From 66b21396fee35b23088070587cc374d83e5f8c0c Mon Sep 17 00:00:00 2001 From: Evgeny Frolov Date: Sat, 10 Mar 2018 18:15:16 +0300 Subject: [PATCH 01/21] fix imports --- polara/recommender/models.py | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/polara/recommender/models.py b/polara/recommender/models.py index 89d2a3f..dd0cb6b 100644 --- a/polara/recommender/models.py +++ b/polara/recommender/models.py @@ -16,8 +16,9 @@ from scipy.sparse import coo_matrix, csr_matrix from scipy.sparse.linalg import svds -from polara.recommender import data, defaults +from polara.recommender import defaults from polara.recommender.evaluation import get_hits, get_relevance_scores, get_ranking_scores +from polara.recommender.evaluation import get_hr_score, get_mrr_score from polara.recommender.evaluation import assemble_scoring_matrices from polara.recommender.utils import array_split, NNZ_MAX from polara.lib.hosvd import tucker_als From c919d1948960a69089814ea83d620b225b2be494 Mon Sep 17 00:00:00 2001 From: Evgeny Frolov Date: Sat, 10 Mar 2018 18:15:49 +0300 Subject: [PATCH 02/21] remove orfan methods --- polara/recommender/models.py | 40 ------------------------------------ 1 file changed, 40 deletions(-) diff --git a/polara/recommender/models.py b/polara/recommender/models.py index dd0cb6b..8c8e9e8 100644 --- a/polara/recommender/models.py +++ b/polara/recommender/models.py @@ -373,46 +373,6 @@ def evaluate(self, method='hits', topk=None, not_rated_penalty=None, on_feedback return scores - def _get_matched_predictions(self): - userid, itemid = self.data.fields.userid, self.data.fields.itemid - holdout_data = self.data.test.holdout[itemid] - holdout_size = self.data.holdout_size - holdout_matrix = holdout_data.values.reshape(-1, holdout_size).astype(np.int64) - - recommendations = self.recommendations #will recalculate if empty - - if recommendations.shape[0] != holdout_matrix.shape[0]: - raise ValueError('Incompatible test and holdout size.') - - matched_predictions = (recommendations[:, :, None] == holdout_matrix[:, None, :]) - return matched_predictions - - - def _get_feedback_data(self, on_level=None): - feedback = self.data.fields.feedback - eval_data = self.data.test.holdout[feedback].values - holdout = self.data.holdout_size - feedback_data = eval_data.reshape(-1, holdout) - - if on_level is not None: - try: - iter(on_level) - except TypeError: - feedback_data = np.ma.masked_not_equal(feedback_data, on_level) - else: - mask_level = np.in1d(feedback_data.ravel(), - on_level, - invert=True).reshape(feedback_data.shape) - feedback_data = np.ma.masked_where(mask_level, feedback_data) - return feedback_data - - - def _get_positive_feedback(self, on_level=None): - feedback_data = self._get_feedback_data(on_level) - positive_feedback = feedback_data >= self.switch_positive - return positive_feedback - - @staticmethod def topsort(a, topk): parted = np.argpartition(a, -topk)[-topk:] From 261aad24bb447f241979e3b041fa08b9e4b05f64 Mon Sep 17 00:00:00 2001 From: Evgeny Frolov Date: Sat, 10 Mar 2018 18:16:31 +0300 Subject: [PATCH 03/21] fix popularity based model --- polara/recommender/models.py | 1 + 1 file changed, 1 insertion(+) diff --git a/polara/recommender/models.py b/polara/recommender/models.py index 8c8e9e8..c21096d 100644 --- a/polara/recommender/models.py +++ b/polara/recommender/models.py @@ -542,6 +542,7 @@ def build(self): itemid = self.data.fields.itemid item_groups = self.data.training.groupby(itemid, sort=True) if self.by_feedback_value: + feedback = self.data.fields.feedback self.items_scores = item_groups[feedback].sum().values else: self.item_scores = item_groups.size().values From 08c8955c0f86948391012184ee437eb925c333f3 Mon Sep 17 00:00:00 2001 From: Evgeny Frolov Date: Sat, 10 Mar 2018 18:24:36 +0300 Subject: [PATCH 04/21] rearrange coffee slice_recommendations code --- polara/recommender/models.py | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/polara/recommender/models.py b/polara/recommender/models.py index c21096d..d6625fc 100644 --- a/polara/recommender/models.py +++ b/polara/recommender/models.py @@ -779,14 +779,15 @@ def get_test_tensor(self, test_data, shape, start, end): def slice_recommendations(self, test_data, shape, start, stop, test_users=None): test_tensor_unfolded, slice_idx = self.get_test_tensor(test_data, shape, start, stop) + v = self.factors[self.data.fields.itemid] + w = self.factors[self.data.fields.feedback] + num_users = stop - start num_items = shape[1] num_fdbks = shape[2] - v = self.factors[self.data.fields.itemid] - w = self.factors[self.data.fields.feedback] # assume that w.shape[1] < v.shape[1] (allows for more efficient calculations) - scores = test_tensor_unfolded.dot(w).reshape(num_users, num_items, w.shape[1]) + scores = test_tensor_unfolded.dot(w).reshape(num_users, num_items, num_fdbks) scores = np.tensordot(scores, v, axes=(1, 0)) scores = np.tensordot(np.tensordot(scores, v, axes=(2, 1)), w, axes=(1, 1)) scores = self.flatten_scores(scores, self.flattener) From 58f4980f0439778e14245ffb9a9cfcd5a97503f2 Mon Sep 17 00:00:00 2001 From: Evgeny Frolov Date: Sat, 10 Mar 2018 19:00:02 +0300 Subject: [PATCH 05/21] fix long type for python 2/3 interop --- polara/lib/similarity.py | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/polara/lib/similarity.py b/polara/lib/similarity.py index 0965394..25288be 100644 --- a/polara/lib/similarity.py +++ b/polara/lib/similarity.py @@ -1,10 +1,16 @@ # python 2/3 interoperability from __future__ import division + try: range = xrange except NameError: pass +try: + long +except NameError: + long = int + import math import types from collections import defaultdict, OrderedDict From d2a03e5853493c6228cf6f6c6efebd3b68664ece Mon Sep 17 00:00:00 2001 From: Evgeny Frolov Date: Sat, 10 Mar 2018 19:05:37 +0300 Subject: [PATCH 06/21] fix imports, decorators, unused variables --- polara/lib/similarity.py | 3 +-- polara/lib/sparse.py | 6 ++++-- polara/recommender/coldstart/data.py | 2 -- polara/recommender/external/mymedialite/mmlwrapper.py | 1 - 4 files changed, 5 insertions(+), 7 deletions(-) diff --git a/polara/lib/similarity.py b/polara/lib/similarity.py index 25288be..2279b27 100644 --- a/polara/lib/similarity.py +++ b/polara/lib/similarity.py @@ -15,7 +15,6 @@ import types from collections import defaultdict, OrderedDict import numpy as np -import pandas as pd from numba import jit import scipy as sp import scipy.sparse @@ -209,7 +208,7 @@ def jaccard_similarity_weighted(F, fill_diagonal=True): shift = 1 if fill_diagonal else 0 data, rows, cols = _jaccard_similarity_weighted_tri(dat, ind, ptr, shift) - S = sp.sparse.coo_matrix((data, (rows, cols)), shape=(F.shape[0],)*2).tocsc() + S = coo_matrix((data, (rows, cols)), shape=(F.shape[0],)*2).tocsc() S += S.T # doubles diagonal values if fill_diagonal is False if fill_diagonal: diff --git a/polara/lib/sparse.py b/polara/lib/sparse.py index 5d31daf..5d38b10 100644 --- a/polara/lib/sparse.py +++ b/polara/lib/sparse.py @@ -6,12 +6,13 @@ import numpy as np import scipy as sp -from scipy import sparse +import scipy.sparse from numba import jit # matvec implementation is based on # http://stackoverflow.com/questions/18595981/improving-performance-of-multiplication-of-scipy-sparse-matrices + @jit(nopython=True, nogil=True) def matvec2dense(m_ptr, m_ind, m_val, v_nnz, v_val, out): l = len(v_nnz) @@ -63,7 +64,8 @@ def csc_matvec(mat_csc, vec, dense_output=True, dtype=None): res.sum_duplicates() # expensive operation return res -jit(nopython=True) + +@jit(nopython=True) def _blockify(ind, ptr, major_dim): # convenient function to compute only diagonal # elements of the product of 2 matrices; diff --git a/polara/recommender/coldstart/data.py b/polara/recommender/coldstart/data.py index c39b731..9038e12 100644 --- a/polara/recommender/coldstart/data.py +++ b/polara/recommender/coldstart/data.py @@ -1,6 +1,5 @@ from collections import namedtuple, defaultdict import numpy as np -import pandas as pd from polara.recommender.data import RecommenderData from polara.lib.similarity import build_indicator_matrix @@ -42,7 +41,6 @@ def prepare(self): def _split_test_index(self): - userid = self.fields.userid itemid = self.fields.itemid item_idx = np.arange(len(self._unique_items)) diff --git a/polara/recommender/external/mymedialite/mmlwrapper.py b/polara/recommender/external/mymedialite/mmlwrapper.py index 3a65f12..1b0d4b2 100644 --- a/polara/recommender/external/mymedialite/mmlwrapper.py +++ b/polara/recommender/external/mymedialite/mmlwrapper.py @@ -76,7 +76,6 @@ def command(self): def _save_to_disk(self): if self.positive_only: feedback = self.data.fields.feedback - userid = self.data.fields.userid pos_ind = self.data.training[feedback] >= self.switch_positive pos_data = self.data.training.loc[pos_ind] pos_data.to_csv(self.train_data_path, index=False, header=False) From 8ffa4111e14da915be7b5d3f1d8560083a941033 Mon Sep 17 00:00:00 2001 From: Evgeny Frolov Date: Sat, 10 Mar 2018 19:06:18 +0300 Subject: [PATCH 07/21] fix bookcrossing data download --- polara/datasets/bookcrossing.py | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/polara/datasets/bookcrossing.py b/polara/datasets/bookcrossing.py index bf0fdf2..a2ef145 100644 --- a/polara/datasets/bookcrossing.py +++ b/polara/datasets/bookcrossing.py @@ -1,5 +1,5 @@ +from io import BytesIO import pandas as pd -from StringIO import StringIO try: from pandas.io.common import ZipFile @@ -13,7 +13,7 @@ def get_bx_data(local_file=None, get_ratings=True, get_users=False, get_books=Fa from requests import get zip_file_url = 'http://www2.informatik.uni-freiburg.de/~cziegler/BX/BX-CSV-Dump.zip' zip_response = get(zip_file_url) - zip_contents = StringIO(zip_response.content) + zip_contents = BytesIO(zip_response.content) else: zip_contents = local_file @@ -26,7 +26,7 @@ def get_bx_data(local_file=None, get_ratings=True, get_users=False, get_books=Fa delimiter = ';' if get_ratings: zdata = zfile.read(zip_file) - ratings = pd.read_csv(StringIO(zdata), sep=delimiter, header=0, engine='c') + ratings = pd.read_csv(BytesIO(zdata), sep=delimiter, header=0, engine='c') if get_users: zip_file = zip_files[zip_files.str.contains('users', flags=2)].iat[0] From 466c2b4e57138b5820af8ea12885595e5a85041c Mon Sep 17 00:00:00 2001 From: Evgeny Frolov Date: Sat, 10 Mar 2018 21:15:18 +0300 Subject: [PATCH 08/21] fix whitespaces, indentation, blank lines --- polara/datasets/movielens.py | 3 +- polara/datasets/netflix.py | 8 +- polara/evaluation/evaluation_engine.py | 4 +- polara/evaluation/plotting.py | 4 +- polara/lib/hosvd.py | 2 + polara/lib/similarity.py | 30 ++-- polara/lib/sparse.py | 3 + polara/recommender/data.py | 160 +++++++++--------- polara/recommender/defaults.py | 2 +- polara/recommender/evaluation.py | 18 +- .../external/graphlab/glwrapper.py | 2 +- polara/recommender/models.py | 88 +++++----- polara/tools/preprocessing.py | 3 +- polara/tools/printing.py | 2 + polara/tools/timing.py | 1 + 15 files changed, 172 insertions(+), 158 deletions(-) diff --git a/polara/datasets/movielens.py b/polara/datasets/movielens.py index ebea0cd..3d8d84e 100644 --- a/polara/datasets/movielens.py +++ b/polara/datasets/movielens.py @@ -6,8 +6,9 @@ except ImportError: from zipfile import ZipFile + def get_movielens_data(local_file=None, get_ratings=True, get_genres=False, - split_genres=True, mdb_mapping=False): + split_genres=True, mdb_mapping=False): '''Downloads movielens data and stores it in pandas dataframe. ''' if not local_file: diff --git a/polara/datasets/netflix.py b/polara/datasets/netflix.py index f095f66..8acc7d5 100644 --- a/polara/datasets/netflix.py +++ b/polara/datasets/netflix.py @@ -17,14 +17,14 @@ def get_netflix_data(gz_file): movie_data.append(df[movieid]) data = pd.concat(movie_data, keys=movie_name) - data = data.reset_index().iloc[:, :3].rename(columns={'level_0':'movieid', - 'level_1':'userid', - 'level_2':'rating'}) + data = data.reset_index().iloc[:, :3].rename(columns={'level_0': 'movieid', + 'level_1': 'userid', + 'level_2': 'rating'}) return data def filter_by_length(data, session_length=20): sz = data.groupby('userid', sort=False).size() valid_users = sz.index[(sz > session_length)] - new_data = data[data.userid.isin(valid_users)] + new_data = data[data.userid.isin(valid_users)] return new_data diff --git a/polara/evaluation/evaluation_engine.py b/polara/evaluation/evaluation_engine.py index 7bbcdcd..482375b 100644 --- a/polara/evaluation/evaluation_engine.py +++ b/polara/evaluation/evaluation_engine.py @@ -37,7 +37,7 @@ def evaluate_models(models, metrics, topk=None): for metric in metrics: model_scores = [] for model in models: - #print('model {}'.format(model.method)) + # print('model {}'.format(model.method)) scores = model.evaluate(method=metric, topk=topk) model_scores.append(scores) metric_scores.append(pd.DataFrame(model_scores, index=[model.method for model in models]).T) @@ -62,7 +62,7 @@ def consolidate(scores, params, metrics): return res -def consolidate_folds(scores, folds, metrics, index_names = ['fold', 'top-n']): +def consolidate_folds(scores, folds, metrics, index_names=['fold', 'top-n']): res = {} for metric in metrics: data = pd.concat([scores[j][metric] for j in folds], keys=folds) diff --git a/polara/evaluation/plotting.py b/polara/evaluation/plotting.py index d83847e..cae01f9 100644 --- a/polara/evaluation/plotting.py +++ b/polara/evaluation/plotting.py @@ -1,7 +1,7 @@ import matplotlib.pyplot as plt -def _plot_pair(scores, keys, titles=None, errors=None, err_alpha = 0.2, figsize=(16, 5), ax=None): +def _plot_pair(scores, keys, titles=None, errors=None, err_alpha=0.2, figsize=(16, 5), ax=None): if not ax: fig, ax = plt.subplots(1, 2, figsize=figsize) show_legend = True @@ -54,7 +54,7 @@ def show_ranking(all_scores, **kwargs): _plot_pair(scores, keys, **kwargs) -def _cross_plot(scores, keys, titles=None, errors=None, err_alpha = 0.2, ROC_middle=False, figsize=(8, 5), limit=None, ax=None): +def _cross_plot(scores, keys, titles=None, errors=None, err_alpha=0.2, ROC_middle=False, figsize=(8, 5), limit=None, ax=None): if not ax: fig = plt.figure(figsize=figsize) ax = fig.gca() diff --git a/polara/lib/hosvd.py b/polara/lib/hosvd.py index 8cb727e..a1bc085 100644 --- a/polara/lib/hosvd.py +++ b/polara/lib/hosvd.py @@ -8,6 +8,7 @@ import numpy as np from numba import jit + @jit(nopython=True, nogil=True) def double_tensordot(idx, val, U, V, new_shape1, new_shape2, ten_mode0, ten_mode1, ten_mode2, res): I = idx.shape[0] @@ -43,6 +44,7 @@ def tensordot2(idx, val, shape, U, V, modes): double_tensordot(idx, val, vU, vV, new_shape[1], new_shape[2], ten_mode0, ten_mode1, ten_mode2, res) return res + def tucker_als(idx, val, shape, core_shape, iters=25, growth_tol=0.01, batch_run=False): ''' The function computes Tucker ALS decomposition of sparse tensor diff --git a/polara/lib/similarity.py b/polara/lib/similarity.py index 2279b27..cbf8ff2 100644 --- a/polara/lib/similarity.py +++ b/polara/lib/similarity.py @@ -41,7 +41,7 @@ def _fix_empty_features(feature_mat): def set_diagonal_values(mat, val=1): - #disable warning when setting diagonal elements of sparse similarity matrix + # disable warning when setting diagonal elements of sparse similarity matrix with warnings.catch_warnings(): warnings.simplefilter('ignore', category=SparseEfficiencyWarning) mat.setdiag(val) @@ -58,7 +58,7 @@ def normalize_binary_features(feature_mat): if feature_mat.format == 'csc': ind = feature_mat.indices.copy() ptr = feature_mat.indptr.copy() - norm_data = 1 / np.sqrt(np.take(sqsum, ind)) + norm_data = 1 / np.sqrt(np.take(sqsum, ind)) normed = csc_matrix((norm_data, ind, ptr), shape=feature_mat.shape) else: norm_data = safe_inverse_root(sqsum.astype(np.float64)) @@ -149,10 +149,10 @@ def cosine_tfidf_similarity(F, fill_diagonal=True): @jit(nopython=True) def _jaccard_similarity_weighted_tri(dat, ind, ptr, shift): z = dat[0] #np.float64 - #need to initialize lists with certain dtype - data = [z,] - cols = [z,] - rows = [z,] + # need to initialize lists with certain dtype + data = [z] + cols = [z] + rows = [z] nrows = len(ptr) - 1 for i in range(nrows): @@ -188,7 +188,7 @@ def _jaccard_similarity_weighted_tri(dat, ind, ptr, shift): max_sum += dat[s] if min_sum: - wjac = min_sum/max_sum + wjac = min_sum / max_sum data.append(wjac) cols.append(i) rows.append(j) @@ -297,10 +297,10 @@ def get_features_data(meta_data, ranking=None, deduplicate=True): ranking = 'linear' if isinstance(ranking, str): - ranking = [ranking,] * len(features) + ranking = [ranking] * len(features) if not isinstance(ranking, dict): - ranking = {k:v for k, v in zip(features, ranking)} + ranking = {k: v for k, v in zip(features, ranking)} for feature in features: feature_data = meta_data[feature] @@ -330,10 +330,10 @@ def get_similarity_data(meta_data, similarity_type='jaccard'): features = meta_data.columns if isinstance(similarity_type, str): - similarity_type = [similarity_type,] * len(features) + similarity_type = [similarity_type] * len(features) if not isinstance(similarity_type, dict): - similarity_type = {k:v for k, v in zip(features, similarity_type)} + similarity_type = {k: v for k, v in zip(features, similarity_type)} similarity_mats = {} for feature in features: @@ -361,16 +361,16 @@ def combine_similarity_data(meta_data, similarity_type='jaccard', weights=None): num_feat = len(features) if isinstance(similarity_type, str): - similarity_type = [similarity_type,] * num_feat + similarity_type = [similarity_type] * num_feat if not isinstance(similarity_type, dict): - similarity_type = {k:v for k, v in zip(features, similarity_type)} + similarity_type = {k: v for k, v in zip(features, similarity_type)} if weights is None: - weights = [1.0/num_feat,] * num_feat + weights = [1.0/num_feat] * num_feat if not isinstance(weights, dict): - weights = {k:v for k, v in zip(features, weights)} + weights = {k: v for k, v in zip(features, weights)} similarity = csc_matrix((meta_data.shape[0],)*2) for feature in features: diff --git a/polara/lib/sparse.py b/polara/lib/sparse.py index 5d38b10..d115c18 100644 --- a/polara/lib/sparse.py +++ b/polara/lib/sparse.py @@ -79,17 +79,20 @@ def _blockify(ind, ptr, major_dim): shift_ind = i * major_dim ind[j] += shift_ind + def row_unblockify(mat, block_size): # only for CSR matrices factor = (mat.indices // block_size) * block_size mat.indices -= factor mat._shape = (mat.shape[0], block_size) + def row_blockify(mat, block_size): # only for CSR matrices _blockify(mat.indices, mat.indptr, block_size) mat._shape = (mat.shape[0], block_size*mat.shape[0]) + def inverse_permutation(p): s = np.empty(p.size, p.dtype) s[p] = np.arange(p.size) diff --git a/polara/recommender/data.py b/polara/recommender/data.py index 6b7f4b7..69ec83a 100644 --- a/polara/recommender/data.py +++ b/polara/recommender/data.py @@ -8,11 +8,13 @@ import numpy as np from polara.recommender import defaults + def random_choice(df, num, random_state): n = df.shape[0] k = min(num, n) return df.iloc[random_state.choice(n, k, replace=False)] + def random_sample(df, frac, random_state): return df.sample(frac=frac, random_state=random_state) @@ -35,7 +37,7 @@ def _get_subscribers(self, event): def subscribe(self, event, callback): subscriber = callback.__self__ - func = callback.__func__ + func = callback.__func__ self._get_subscribers(event).setdefault(subscriber, set()).add(func) def unsubscribe(self, event, subscriber): @@ -99,8 +101,8 @@ def __init__(self, data, userid, itemid, feedback, custom_order=None, seed=None) self._data = data if data.duplicated(subset=fields).any(): - #unstable in pandas v. 17.0, only works in <> v.17.0 - #rely on deduplicated data in many places - makes data processing more efficient + # unstable in pandas v. 17.0, only works in <> v.17.0 + # rely on deduplicated data in many places - makes data processing more efficient raise NotImplementedError('Data has duplicate values') self._custom_order = custom_order @@ -110,16 +112,16 @@ def __init__(self, data, userid, itemid, feedback, custom_order=None, seed=None) self.index = self.index._make([None]*len(self._std_fields)) self._set_defaults() - self._change_properties = set() #container for changed properties + self._change_properties = set() # container for changed properties # depending on config. For ex., test_fold - full_update, # TODO seed may also lead to either full_update or test_update # random_holdout - test_update. Need to implement checks - self.seed = seed #use with permute_tops, random_choice + self.seed = seed # use with permute_tops, random_choice self.verify_sessions_length_distribution = True - self.ensure_consistency = True # drop test entities if not present in training - self.build_index = True # reindex data, avoid gaps in user and item index + self.ensure_consistency = True # drop test entities if not present in training + self.build_index = True # reindex data, avoid gaps in user and item index self._test_selector = None - self._state = None # None or 1 of {'_': 1, 'H': 11, '|': 2, 'd': 3, 'T': 4} + self._state = None # None or 1 of {'_': 1, 'H': 11, '|': 2, 'd': 3, 'T': 4} self._last_update_rule = None self.on_change_event = 'on_change' @@ -138,7 +140,7 @@ def unsubscribe(self, event, model): def _set_defaults(self, params=None): - #[1:] omits undersacores in properties names + # [1:] omits undersacores in properties names params = params or [prop[1:] for prop in self._config] config_vals = defaults.get_config(params) for name, value in config_vals.items(): @@ -147,8 +149,8 @@ def _set_defaults(self, params=None): def get_configuration(self): - #[1:] omits undersacores in properties names, i.e. uses external name - #in that case it prints worning if change is pending + # [1:] omits undersacores in properties names, i.e. uses external name + # in that case it prints worning if change is pending config = {attr[1:]: getattr(self, attr[1:]) for attr in self._config} return config @@ -160,7 +162,7 @@ def test(self): @property def training(self): - self.update() # both _test and _training attributes appear simultaneously + self.update() # both _test and _training attributes appear simultaneously return self._training @@ -196,19 +198,19 @@ def prepare(self): self._try_reindex_training_data() if update_rule['full_update'] or update_rule['test_update']: - self._try_drop_unseen_test_items() # unseen = not present in training data - self._try_drop_unseen_test_users() # unseen = not present in training data - self._try_drop_invalid_test_users() # with too few items and/or if inconsistent between testset and holdout - self._try_reindex_test_data() # either assign known index, or (if testing for unseen users) reindex + self._try_drop_unseen_test_items() # unseen = not present in training data + self._try_drop_unseen_test_users() # unseen = not present in training data + self._try_drop_invalid_test_users() # with too few items and/or if inconsistent between testset and holdout + self._try_reindex_test_data() # either assign known index, or (if testing for unseen users) reindex self._try_sort_test_data() if self.verbose: print('Done.') def prepare_training_only(self): - self.holdout_size = 0 # do not form holdout - self.test_ratio = 0 # do not form testset - self.warm_start = False # required for correct state transition handling + self.holdout_size = 0 # do not form holdout + self.test_ratio = 0 # do not form testset + self.warm_start = False # required for correct state transition handling self.prepare() @@ -222,7 +224,7 @@ def _validate_config(self): if self._test_ratio: max_fold = 1.0 / self._test_ratio - if self._test_fold > max_fold: + if self._test_fold > max_fold: raise ValueError('Test fold value cannot be greater than {}'.format(max_fold)) @@ -244,7 +246,7 @@ def _check_state_transition(self): update_rule = defaultdict(bool) new_state = last_state - if unseen_usr_change: # unseen_test_users is reserved for state 4 only! + if unseen_usr_change: # unseen_test_users is reserved for state 4 only! if test_unseen: new_state = 4 if (last_state == 11) and not test_data_change: @@ -266,73 +268,73 @@ def _check_state_transition(self): new_state = 2 else: new_state = 3 - else: # this assumes that warm_start is consistent with current state! - if last_state == 1: # hsz = 0, trt = 0, usn = False - if holdout_sz_change: # hsz > 0 + else: # this assumes that warm_start is consistent with current state! + if last_state == 1: # hsz = 0, trt = 0, usn = False + if holdout_sz_change: # hsz > 0 new_state = 3 if test_ratio_change else 2 update_rule['full_update'] = True - elif test_ratio_change: # hsz = 0, trt > 0 + elif test_ratio_change: # hsz = 0, trt > 0 new_state = 11 update_rule['full_update'] = True - elif last_state == 11: # hsz = 0, trt > 0, usn = False - if holdout_sz_change: # hsz > 0 + elif last_state == 11: # hsz = 0, trt > 0, usn = False + if holdout_sz_change: # hsz > 0 new_state = 2 if empty_testset else 3 update_rule['full_update'] = True - elif test_data_change: # hsz = 0 - if empty_testset: # hsz = 0, trt = 0 + elif test_data_change: # hsz = 0 + if empty_testset: # hsz = 0, trt = 0 new_state = 1 update_rule['full_update'] = True - elif last_state == 2: # hsz > 0, trt = 0, usn = False - if test_ratio_change: # trt > 0 + elif last_state == 2: # hsz > 0, trt = 0, usn = False + if test_ratio_change: # trt > 0 new_state = 11 if empty_holdout else 3 update_rule['full_update'] = True - elif any_holdout_change: # trt = 0 - if empty_holdout: # hsz = 0 + elif any_holdout_change: # trt = 0 + if empty_holdout: # hsz = 0 new_state = 1 update_rule['full_update'] = True - elif last_state == 3: # hsz > 0, trt > 0, usn = False + elif last_state == 3: # hsz > 0, trt > 0, usn = False if test_data_change or any_holdout_change: if empty_holdout: new_state = 1 if empty_testset else 11 - elif empty_testset: # hsz > 0, trt = 0 + elif empty_testset: # hsz > 0, trt = 0 new_state = 2 update_rule['full_update'] = True - elif last_state == 4: # hsz > 0, trt > 0, usn = True + elif last_state == 4: # hsz > 0, trt > 0, usn = True if any_holdout_change: if empty_holdout: if test_data_change: new_state = 1 if empty_testset else 11 update_rule['full_update'] = True - else: # hsz = 0, trt > 0 + else: # hsz = 0, trt > 0 new_state = 11 update_rule['test_update'] = True - else: # hsz > 0 + else: # hsz > 0 if test_data_change: - if empty_testset: # hsz > 0, trt = 0 + if empty_testset: # hsz > 0, trt = 0 new_state = 2 update_rule['full_update'] = True - else: # including test_sample_change + else: # including test_sample_change update_rule['test_update'] = True - else: # hsz > 0 + else: # hsz > 0 if test_data_change: - if empty_testset: # hsz > 0, trt = 0 + if empty_testset: # hsz > 0, trt = 0 new_state = 2 update_rule['full_update'] = True elif test_sample_change: update_rule['test_update'] = True - else: # initial state + else: # initial state if empty_holdout: new_state = 1 if empty_testset else 11 else: - if empty_testset: # hsz > 0, trt = 0 + if empty_testset: # hsz > 0, trt = 0 new_state = 2 - else: # hsz > 0, trt > 0 + else: # hsz > 0, trt > 0 new_state = 4 if test_unseen else 3 update_rule['full_update'] = True @@ -355,33 +357,33 @@ def _split_data(self): if self._test_ratio > 0: if full_update: test_split = self._split_test_index() - else: #test_update + else: # test_update test_split = self._test_split if self._holdout_size == 0: # state 11 testset = holdout = None train_split = ~test_split - else: # state 3 or state 4 + else: # state 3 or state 4 # NOTE holdout_size = None is also here; this can be used in # subclasses like ItemColdStartData to preprocess data properly # in that case _sample_holdout must be modified accordingly holdout = self._sample_holdout(test_split) - if self._warm_start: # state 4 + if self._warm_start: # state 4 testset = self._sample_testset(test_split, holdout.index) train_split = ~test_split - else: # state 3 - testset = None # will be computed if test data is requested + else: # state 3 + testset = None # will be computed if test data is requested train_split = ~self._data.index.isin(holdout.index) - else: # test_ratio == 0 - testset = None # will be computed if test data is requested + else: # test_ratio == 0 + testset = None # will be computed if test data is requested test_split = slice(None) - if self._holdout_size >= 1: # state 2, sample holdout data per each user + if self._holdout_size >= 1: # state 2, sample holdout data per each user holdout = self._sample_holdout(test_split) - elif self._holdout_size > 0: # state 2, special case - sample whole data at once + elif self._holdout_size > 0: # state 2, special case - sample whole data at once random_state = np.random.RandomState(self.seed) holdout = self._data.sample(frac=self._holdout_size, random_state=random_state) - else: # state 1 + else: # state 1 holdout = None train_split = slice(None) if holdout is None else ~self._data.index.isin(holdout.index) @@ -410,7 +412,7 @@ def _split_test_index(self): def _get_sessions_info(self): userid = self.fields.userid - user_sessions = self._data.groupby(userid, sort=True) #KEEP TRUE HERE! + user_sessions = self._data.groupby(userid, sort=True) # KEEP TRUE HERE! # if False than long sessions idx are prevalent in the beginning => non-equal size folds # this effect is taken into account with help of is_not_uniform function # example (run several times to see a pattern): @@ -478,8 +480,8 @@ def _try_drop_unseen_test_users(self): def _try_drop_invalid_test_users(self): if self.holdout_size >= 1: # TODO remove that, when new evaluation arrives - self._filter_short_sessions() # ensure holdout conforms the holdout_size attribute - self._align_test_users() # ensure the same users are in both testset and holdout + self._filter_short_sessions() # ensure holdout conforms the holdout_size attribute + self._align_test_users() # ensure the same users are in both testset and holdout def _try_reindex_test_data(self): self._assign_test_items_index() @@ -600,8 +602,8 @@ def _filter_unseen_entity(self, entity, dataset, label): if not seen_data.all(): n_unseen_entities = dataset.loc[~seen_data, entity].nunique() dataset.query('{} in @seen_entities'.format(entity), inplace=True) - #unseen_index = dataset.index[unseen_entities] - #dataset.drop(unseen_index, inplace=True) + # unseen_index = dataset.index[unseen_entities] + # dataset.drop(unseen_index, inplace=True) if self.verbose: UNSEEN = 'not in the training data' msg = '{} unique {}\'s within {} {} interactions were filtered. Reason: {}.' @@ -658,19 +660,19 @@ def _sample_holdout(self, test_split, group_id=None): group_id = group_id or self.fields.userid grouper = selector.groupby(self._data[group_id], sort=False) - if self._random_holdout: #randomly sample data for evaluation + if self._random_holdout: # randomly sample data for evaluation random_state = np.random.RandomState(self.seed) - if self._holdout_size >= 1: # pick at most _holdout_size elements + if self._holdout_size >= 1: # pick at most _holdout_size elements holdout = grouper.apply(random_choice, self._holdout_size, random_state) else: holdout = grouper.apply(random_sample, self._holdout_size, random_state) - elif self._negative_prediction: #try to holdout negative only examples - if self._holdout_size >= 1: # pick at most _holdout_size elements + elif self._negative_prediction: # try to holdout negative only examples + if self._holdout_size >= 1: # pick at most _holdout_size elements holdout = grouper.nsmallest(self._holdout_size, keep='last') else: raise NotImplementedError - else: #standard top-score prediction mode - if self._holdout_size >= 1: # pick at most _holdout_size elements + else: # standard top-score prediction mode + if self._holdout_size >= 1: # pick at most _holdout_size elements holdout = grouper.nlargest(self._holdout_size, keep='last') else: raise NotImplementedError @@ -687,14 +689,14 @@ def _sample_testset(self, test_split, holdout_index): return data userid = self.fields.userid - if test_sample > 0: # sample at most test_sample items + if test_sample > 0: # sample at most test_sample items random_state = np.random.RandomState(self.seed) sampled = (data.groupby(userid, sort=False, group_keys=False) - .apply(random_choice, test_sample, random_state)) - else: # sample at most test_sample items with the worst feedback from user + .apply(random_choice, test_sample, random_state)) + else: # sample at most test_sample items with the worst feedback from user feedback = self.fields.feedback idx = (data.groupby(userid, sort=False)[feedback] - .nsmallest(-test_sample).index.get_level_values(1)) + .nsmallest(-test_sample).index.get_level_values(1)) sampled = data.loc[idx] return sampled @@ -730,7 +732,7 @@ def _recover_testset(self, update_data=False): testset = self.training else: testset = (self.training.query('{} in @test_users'.format(userid)) - .sort_values(userid)) + .sort_values(userid)) if update_data: self._test = self._test._replace(testset=testset) return testset @@ -767,7 +769,7 @@ def get_test_shape(self, tensor_mode=False): userid = self.fields.userid if self.test.holdout is None: num_users = self.test.testset[userid].nunique() - #TODO make it a property + # TODO make it a property else: num_users = self.test.holdout[userid].nunique() @@ -817,14 +819,14 @@ def set_test_data(self, testset=None, holdout=None, warm_start=False, test_users self._notify(self.on_update_event) self._change_properties.clear() - if (testset is None) and (holdout is None): return # allows to cleanup data + if (testset is None) and (holdout is None): return # allows to cleanup data - if ensure_consistency: # allows to disable self.ensure_consistency without actually changing it - self._try_drop_unseen_test_items() # unseen = not present in training data - self._try_drop_unseen_test_users() # unseen = not present in training data - self._try_drop_invalid_test_users() # inconsistent between testset and holdout + if ensure_consistency: # allows to disable self.ensure_consistency without actually changing it + self._try_drop_unseen_test_items() # unseen = not present in training data + self._try_drop_unseen_test_users() # unseen = not present in training data + self._try_drop_invalid_test_users() # inconsistent between testset and holdout if reindex: - self._try_reindex_test_data() # either assign known index, or reindex (if warm_start) + self._try_reindex_test_data() # either assign known index, or reindex (if warm_start) self._try_sort_test_data() @@ -896,7 +898,7 @@ def _get_long_tail(self): tail_idx = None if self.head_items_frac: - self.head_feedback_frac = None # could in principle calculate real value instead + self.head_feedback_frac = None # could in principle calculate real value instead items_frac = np.arange(1, len(popularity)+1) / len(popularity) tail_idx = items_frac > self.head_items_frac diff --git a/polara/recommender/defaults.py b/polara/recommender/defaults.py index d1a1b25..b712fbc 100644 --- a/polara/recommender/defaults.py +++ b/polara/recommender/defaults.py @@ -42,5 +42,5 @@ def get_config(params): this = sys.modules[__name__] - config = {param:getattr(this, param) for param in params} + config = {param: getattr(this, param) for param in params} return config diff --git a/polara/recommender/evaluation.py b/polara/recommender/evaluation.py index 0174713..928cb3f 100644 --- a/polara/recommender/evaluation.py +++ b/polara/recommender/evaluation.py @@ -94,13 +94,13 @@ def assemble_scoring_matrices(recommendations, eval_data, key, target, is_positi def get_hr_score(hits_rank): 'Hit-Rate score' hr = hits_rank.getnnz(axis=1).mean() - return namedtuple('Relevance', ['hr',])._make([hr,]) + return namedtuple('Relevance', ['hr'])._make([hr]) def get_mrr_score(hits_rank): 'Mean Reciprocal Rank score' mrr = hits_rank.power(-1, 'f8').max(axis=1).mean() - return namedtuple('Ranking', ['mrr',])._make([mrr,]) + return namedtuple('Ranking', ['mrr'])._make([mrr]) def get_ndcr_discounts(rank_matrix, eval_matrix, topn): @@ -111,8 +111,8 @@ def get_ndcr_discounts(rank_matrix, eval_matrix, topn): relevance_per_key = np.array_split(eval_matrix.data, eval_matrix.indptr[1:-1]) target_id_per_key = np.array_split(eval_matrix.indices, eval_matrix.indptr[1:-1]) - #ideal_indices = [np.argsort(rel)[:-(topn+1):-1] for rel in relevance_per_key] - #idx = np.arange(2, topn+2) + # ideal_indices = [np.argsort(rel)[:-(topn+1):-1] for rel in relevance_per_key] + # idx = np.arange(2, topn+2) ideal_indices = [np.argsort(rel)[::-1] for rel in relevance_per_key] idx = np.arange(2, eval_matrix.getnnz(axis=1).max()+2) data = np.concatenate([np.reciprocal(np.log2(idx[:len(i)], dtype='f8')) for i in ideal_indices]) @@ -183,16 +183,16 @@ def get_hits(rank_matrix, hits_rank, miss_rank, eval_matrix, eval_matrix_hits, e hits = namedtuple('Hits', ['true_positive', 'false_positive', 'true_negative', 'false_negative']) hits = hits._make(get_relevance_data(rank_matrix, hits_rank, miss_rank, - eval_matrix, eval_matrix_hits, eval_matrix_miss, - not_rated_penalty, False)) + eval_matrix, eval_matrix_hits, eval_matrix_miss, + not_rated_penalty, False)) return hits def get_relevance_scores(rank_matrix, hits_rank, miss_rank, eval_matrix, eval_matrix_hits, eval_matrix_miss, not_rated_penalty=None): [true_positive, false_positive, true_negative, false_negative] = get_relevance_data(rank_matrix, hits_rank, miss_rank, - eval_matrix, eval_matrix_hits, eval_matrix_miss, - not_rated_penalty, True) + eval_matrix, eval_matrix_hits, eval_matrix_miss, + not_rated_penalty, True) with np.errstate(invalid='ignore'): # true positive rate @@ -210,7 +210,7 @@ def get_relevance_scores(rank_matrix, hits_rank, miss_rank, eval_matrix, eval_ma fallout = specifity = None n_keys = hits_rank.shape[0] - #average over all users + # average over all users precision = np.nansum(precision) / n_keys recall = np.nansum(recall) / n_keys miss_rate = np.nansum(miss_rate) / n_keys diff --git a/polara/recommender/external/graphlab/glwrapper.py b/polara/recommender/external/graphlab/glwrapper.py index fdf59f8..ada8a26 100644 --- a/polara/recommender/external/graphlab/glwrapper.py +++ b/polara/recommender/external/graphlab/glwrapper.py @@ -177,7 +177,7 @@ def get_recommendations(self): unseen_items_idx = np.arange(lower_index, upper_index) new_item_data = gl.SFrame(self.item_side_info.loc[cold_items_index] .reset_index() - .assign(**{itemid:unseen_items_idx})) + .assign(**{itemid: unseen_items_idx})) repr_users = self.data.representative_users.new.values observation_idx = [a.flat for a in np.broadcast_arrays(repr_users, unseen_items_idx[:, None])] diff --git a/polara/recommender/models.py b/polara/recommender/models.py index d6625fc..fa4a247 100644 --- a/polara/recommender/models.py +++ b/polara/recommender/models.py @@ -45,7 +45,7 @@ def wrapper(self, *args, **kwargs): def with_metaclass(mcls): # this is used to ensure python 2/3 interoperablity, taken from: - #https://stackoverflow.com/questions/22409430/portable-meta-class-between-python2-and-python3 + # https://stackoverflow.com/questions/22409430/portable-meta-class-between-python2-and-python3 def decorator(cls): body = vars(cls).copy() # clean out class body @@ -69,7 +69,7 @@ def __new__(meta, name, bases, clsdict): @with_metaclass(MetaModel) class RecommenderModel(object): _config = ('topk', 'filter_seen', 'switch_positive', 'verify_integrity') - _pad_const = -1 # used for sparse data + _pad_const = -1 # used for sparse data def __init__(self, recommender_data, switch_positive=None): @@ -78,13 +78,13 @@ def __init__(self, recommender_data, switch_positive=None): self.method = 'ABC' self._topk = get_default('topk') - self.filter_seen = get_default('filter_seen') + self.filter_seen = get_default('filter_seen') # `switch_positive` can be used by other models during construction process # (e.g. mymedialite wrapper or any other implicit model); hence, it's # better to make it a model attribute, not a simple evaluation argument # (in contrast to `on_feedback_level` argument of self.evaluate) - self.switch_positive = switch_positive or get_default('switch_positive') - self.verify_integrity = get_default('verify_integrity') + self.switch_positive = switch_positive or get_default('switch_positive') + self.verify_integrity = get_default('verify_integrity') # TODO sorting in data must be by self._key, also need to change get_test_data self._key = self.data.fields.userid @@ -122,9 +122,9 @@ def topk(self): @topk.setter def topk(self, new_value): - #support rolling back scenarion for @k calculations + # support rolling back scenarion for @k calculations if (self._recommendations is not None) and (new_value > self._recommendations.shape[1]): - self._recommendations = None #if topk is too high - recalculate recommendations + self._recommendations = None # if topk is too high - recalculate recommendations self._topk = new_value @@ -159,8 +159,8 @@ def get_test_matrix(self, test_data=None, shape=None, user_slice=None): user_coo, item_coo, fdbk_coo = coo_data num_items = shape[1] test_matrix = csr_matrix((fdbk_coo, (user_coo, item_coo)), - shape=(num_users, num_items), - dtype=fdbk_coo.dtype) + shape=(num_users, num_items), + dtype=fdbk_coo.dtype) return test_matrix, coo_data @@ -188,12 +188,12 @@ def _get_test_data(self): idx_diff = np.diff(user_idx) # TODO sorting by self._key - assert (idx_diff >= 0).all() # calculations assume testset is sorted by users! + assert (idx_diff >= 0).all() # calculations assume testset is sorted by users! # TODO only required when testset consists of known users - if (idx_diff>1).any() or (user_idx.min() != 0): # check index monotonicity + if (idx_diff > 1).any() or (user_idx.min() != 0): # check index monotonicity test_users = user_idx[np.r_[0, np.where(idx_diff)[0]+1]] - user_idx = np.r_[0, np.cumsum(idx_diff>0)].astype(user_idx.dtype) + user_idx = np.r_[0, np.cumsum(idx_diff > 0)].astype(user_idx.dtype) else: test_users = np.arange(test_shape[0]) @@ -205,7 +205,7 @@ def _get_test_data(self): def _slice_test_data(self, test_data, start, stop): user_coo, item_coo, fdbk_coo = test_data - slicer = (user_coo>=start) & (user_coo= start) & (user_coo < stop) # always slice over users only user_slice_coo = user_coo[slicer] - start item_slice_coo = item_coo[slicer] @@ -251,9 +251,9 @@ def _make_user(self, user_info): # need to convert itemid's to internal representation # conversion is not required for feedback (it's made in *to_coo functions, if needed) items_data = item_index.set_index('old').loc[items_data, 'new'].values - user_data = pd.DataFrame({userid:[0]*len(items_data), - itemid:items_data, - feedback:feedback_data}) + user_data = pd.DataFrame({userid: [0]*len(items_data), + itemid: items_data, + feedback: feedback_data}) return user_data @@ -277,7 +277,7 @@ def show_recommendations(self, user_info, topk=None): _topk = self.topk self.topk = topk or _topk # takes care of both sparse and dense recommendation lists - top_recs = self.get_topk_elements(scores).squeeze() #remove singleton + top_recs = self.get_topk_elements(scores).squeeze() # remove singleton self.topk = _topk try: @@ -285,7 +285,7 @@ def show_recommendations(self, user_info, topk=None): except AttributeError: item_index = self.data.index.itemid - seen_idx = seen_idx[1] # only items idx + seen_idx = seen_idx[1] # only items idx # covert back to external representation item_idx_map = item_index.set_index('new') top_recs = item_idx_map.loc[top_recs, 'old'].values @@ -326,10 +326,10 @@ def get_recommendations(self): def evaluate(self, method='hits', topk=None, not_rated_penalty=None, on_feedback_level=None): feedback = self.data.fields.feedback if int(topk or 0) > self.topk: - self.topk = topk # will also flush old recommendations + self.topk = topk # will also flush old recommendations # support rolling back scenario for @k calculations - recommendations = self.recommendations[:, :topk] # will recalculate if empty + recommendations = self.recommendations[:, :topk] # will recalculate if empty eval_data = self.data.test.holdout if self.switch_positive is None: @@ -352,7 +352,7 @@ def evaluate(self, method='hits', topk=None, not_rated_penalty=None, on_feedback self._key, self._target, is_positive, feedback=feedback) - if method == 'relevance': # no need for feedback + if method == 'relevance': # no need for feedback if self.data.holdout_size == 1: scores = get_hr_score(scoring_data[1]) else: @@ -362,11 +362,11 @@ def evaluate(self, method='hits', topk=None, not_rated_penalty=None, on_feedback scores = get_mrr_score(scoring_data[1]) else: ndcg_alternative = get_default('ndcg_alternative') - topk = recommendations.shape[1] # handle topk=None case + topk = recommendations.shape[1] # handle topk=None case # topk has to be passed explicitly, otherwise it's unclear how to # estimate ideal ranking for NDCG and NDCL metrics in get_ndcr_discounts scores = get_ranking_scores(*scoring_data, switch_positive=self.switch_positive, topk=topk, alternative=ndcg_alternative) - elif method == 'hits': # no need for feedback + elif method == 'hits': # no need for feedback scores = get_hits(*scoring_data, not_rated_penalty=not_rated_penalty) else: raise NotImplementedError @@ -385,9 +385,9 @@ def downvote_seen_items(recs, idx_seen): # results (comparing to the same method but with "densified" scores matrix) # models with sparse scores can alleviate that by extending recommendations # list with most popular items or items generated by a more sophisticated logic - idx_seen = idx_seen[:2] # need only users and items + idx_seen = idx_seen[:2] # need only users and items if sp.sparse.issparse(recs): - ind_data = np.ones(len(idx_seen[0]), dtype=np.bool) # indicator + ind_data = np.ones(len(idx_seen[0]), dtype=np.bool) # indicator seen = coo_matrix((ind_data, idx_seen), shape=recs.shape, copy=False) seen_recs = recs.multiply(seen) # In the sparse case it's impossible to downvote seen items scores @@ -416,6 +416,7 @@ def get_topk_elements(self, scores): # can do this by adding -1's to the right, however: # this relies on the fact that there are no -1's in evaluation matrix # NOTE need to ensure that this is always true + def topscore(x, k): data = x.data.values cols = x.cols.values @@ -423,7 +424,8 @@ def topscore(x, k): if k >= nnz: cols_sorted = cols[np.argsort(-data)] # need to pad values to conform with evaluation matrix shape - res = np.pad(cols_sorted, (0, k-nnz), 'constant', constant_values=self._pad_const) + res = np.pad(cols_sorted, (0, k-nnz), + 'constant', constant_values=self._pad_const) else: # TODO verify, that even if k is relatively small, then # argpartition doesn't add too much overhead? @@ -513,7 +515,7 @@ def get_recommendations(self): if self.method == 'mostpopular': items_scores = self.data.training.groupby(itemid, sort=True).size().values - #scores = np.lib.stride_tricks.as_strided(items_scores, (num_users, items_scores.size), (0, items_scores.itemsize)) + # scores = np.lib.stride_tricks.as_strided(items_scores, (num_users, items_scores.size), (0, items_scores.itemsize)) scores = np.repeat(items_scores[None, :], num_users, axis=0) elif self.method == 'random': num_items = self.data.training[itemid].max() + 1 @@ -525,10 +527,10 @@ def get_recommendations(self): raise NotImplementedError if self.filter_seen: - #prevent seen items from appearing in recommendations + # prevent seen items from appearing in recommendations self.downvote_seen_items(scores, test_idx) - top_recs = self.get_topk_elements(scores) + top_recs = self.get_topk_elements(scores) return top_recs @@ -579,7 +581,7 @@ class CooccurrenceModel(RecommenderModel): def __init__(self, *args, **kwargs): super(CooccurrenceModel, self).__init__(*args, **kwargs) - self.method = 'item-to-item' #pick some meaningful name + self.method = 'item-to-item' # pick some meaningful name self.implicit = True self.dense_output = False @@ -591,19 +593,19 @@ def build(self): user_item_matrix.data = np.sign(user_item_matrix.data) with Timer(self.method, verbose=self.verbose): - i2i_matrix = user_item_matrix.T.dot(user_item_matrix) # gives CSC format - i2i_matrix.setdiag(0) #exclude "self-links" + i2i_matrix = user_item_matrix.T.dot(user_item_matrix) # gives CSC format + i2i_matrix.setdiag(0) # exclude "self-links" i2i_matrix.eliminate_zeros() self._i2i_matrix = i2i_matrix def _sparse_dot(self, tst_mat, i2i_mat): - # scipy always returns sparse result, even if dot product is actually dense - # this function offers solution to this problem - # it also takes care on sparse result w.r.t. to further processing - if self.dense_output: # calculate dense result directly - # TODO implement matmat multiplication instead of iteration with matvec + # scipy always returns sparse result, even if dot product is dense + # this function offers solution to this problem + # it also takes care on sparse result w.r.t. to further processing + if self.dense_output: # calculate dense result directly + # TODO matmat multiplication instead of iteration with matvec res_type = np.result_type(i2i_mat.dtype, tst_mat.dtype) scores = np.empty((tst_mat.shape[0], i2i_mat.shape[1]), dtype=res_type) for i in range(tst_mat.shape[0]): @@ -625,7 +627,7 @@ def slice_recommendations(self, test_data, shape, start, stop, test_user=None): # recommendations, as vector of shape (1, N) in CSC format is inefficient if self.implicit: - test_matrix.data = np.sign(test_matrix.data) + test_matrix.data = np.sign(test_matrix.data) scores = self._sparse_dot(test_matrix, self._i2i_matrix) return scores, slice_data @@ -749,10 +751,10 @@ def build(self): with Timer(self.method, verbose=self.verbose): users_factors, items_factors, feedback_factors, core = \ - tucker_als(idx, val, shp, self.mlrank, - growth_tol=self.growth_tol, - iters = self.num_iters, - batch_run=not self.show_output) + tucker_als(idx, val, shp, self.mlrank, + growth_tol=self.growth_tol, + iters=self.num_iters, + batch_run=not self.show_output) self.factors[self.data.fields.userid] = users_factors self.factors[self.data.fields.itemid] = items_factors @@ -807,7 +809,7 @@ def get_holdout_slice(self, start, stop): def predict_feedback(self): flattener_old = self.flattener - self.flattener = 'argmax' #this will be applied along feedback axis + self.flattener = 'argmax' # this will be applied along feedback axis feedback_idx = self.data.index.feedback.set_index('new') test_data, test_shape, _ = self._get_test_data() diff --git a/polara/tools/preprocessing.py b/polara/tools/preprocessing.py index 8a37875..9605da0 100644 --- a/polara/tools/preprocessing.py +++ b/polara/tools/preprocessing.py @@ -1,6 +1,7 @@ # python 2/3 interoperability from __future__ import print_function + def filter_sessions_by_length(data, session_label='userid', min_session_length=3): """Filters users with insufficient number of items""" if data.duplicated().any(): @@ -10,7 +11,7 @@ def filter_sessions_by_length(data, session_label='userid', min_session_length=3 has_valid_session_length = sz >= min_session_length if not has_valid_session_length.all(): valid_sessions = sz.index[has_valid_session_length] - new_data = data[data[session_label].isin(valid_sessions)].copy() + new_data = data[data[session_label].isin(valid_sessions)].copy() print('Sessions are filtered by length') else: new_data = data diff --git a/polara/tools/printing.py b/polara/tools/printing.py index 3e86de3..a3e90aa 100644 --- a/polara/tools/printing.py +++ b/polara/tools/printing.py @@ -2,6 +2,7 @@ from contextlib import contextmanager import sys, os + def print_frames(dataframes): if not isinstance(dataframes, tuple): return dataframes @@ -18,6 +19,7 @@ def print_frames(dataframes): return HTML(table) + # from http://thesmithfam.org/blog/2012/10/25/temporarily-suppress-console-output-in-python/# @contextmanager def suppress_stdout(): diff --git a/polara/tools/timing.py b/polara/tools/timing.py index a9f5eb1..bcc4f8b 100644 --- a/polara/tools/timing.py +++ b/polara/tools/timing.py @@ -1,5 +1,6 @@ from timeit import default_timer as timer + class Timer(): def __init__(self, model_name='Model', verbose=True, msg=None): self.model_name = model_name From fcd0f2e179073a7c60ab49c3495ccd99e0ea8913 Mon Sep 17 00:00:00 2001 From: Evgeny Frolov Date: Sat, 10 Mar 2018 21:23:04 +0300 Subject: [PATCH 09/21] fix scipy.sparse imports --- polara/lib/similarity.py | 1 - polara/lib/sparse.py | 5 ++--- polara/recommender/models.py | 1 - 3 files changed, 2 insertions(+), 5 deletions(-) diff --git a/polara/lib/similarity.py b/polara/lib/similarity.py index cbf8ff2..4f8dcc4 100644 --- a/polara/lib/similarity.py +++ b/polara/lib/similarity.py @@ -17,7 +17,6 @@ import numpy as np from numba import jit import scipy as sp -import scipy.sparse from scipy.sparse import csc_matrix, csr_matrix, coo_matrix, SparseEfficiencyWarning import warnings diff --git a/polara/lib/sparse.py b/polara/lib/sparse.py index d115c18..11bde63 100644 --- a/polara/lib/sparse.py +++ b/polara/lib/sparse.py @@ -5,8 +5,7 @@ pass import numpy as np -import scipy as sp -import scipy.sparse +from scipy.sparse import csr_matrix from numba import jit # matvec implementation is based on @@ -60,7 +59,7 @@ def csc_matvec(mat_csc, vec, dense_output=True, dtype=None): indices = np.empty((n,), dtype=np.intp) indptr = np.array([0, n], dtype=np.intp) matvec2sparse(m_ptr, m_ind, m_val, v_nnz, v_val, sizes, indices, data) - res = sp.sparse.csr_matrix((data, indices, indptr), shape=(1, mat_csc.shape[0]), dtype=res_dtype) + res = csr_matrix((data, indices, indptr), shape=(1, mat_csc.shape[0]), dtype=res_dtype) res.sum_duplicates() # expensive operation return res diff --git a/polara/recommender/models.py b/polara/recommender/models.py index fa4a247..cb35e88 100644 --- a/polara/recommender/models.py +++ b/polara/recommender/models.py @@ -12,7 +12,6 @@ import pandas as pd import numpy as np import scipy as sp -import scipy.sparse from scipy.sparse import coo_matrix, csr_matrix from scipy.sparse.linalg import svds From 863083097b55ec869c618f91f34f58ac3aaa82f6 Mon Sep 17 00:00:00 2001 From: Evgeny Frolov Date: Sun, 11 Mar 2018 13:20:31 +0300 Subject: [PATCH 10/21] fix indentation --- polara/evaluation/plotting.py | 2 +- polara/lib/hosvd.py | 6 +++--- polara/recommender/coldstart/data.py | 2 +- polara/recommender/data.py | 4 ++-- polara/recommender/external/mymedialite/mmlwrapper.py | 8 ++++---- polara/recommender/models.py | 4 ++-- 6 files changed, 13 insertions(+), 13 deletions(-) diff --git a/polara/evaluation/plotting.py b/polara/evaluation/plotting.py index cae01f9..3bc3123 100644 --- a/polara/evaluation/plotting.py +++ b/polara/evaluation/plotting.py @@ -15,7 +15,7 @@ def _plot_pair(scores, keys, titles=None, errors=None, err_alpha=0.2, figsize=(1 scores[right].plot(ax=ax[1], legend=False) if show_legend: - plt.legend(loc='center left', bbox_to_anchor=(1.0, 0.5)) + plt.legend(loc='center left', bbox_to_anchor=(1.0, 0.5)) if errors is not None: errG = errors[left] diff --git a/polara/lib/hosvd.py b/polara/lib/hosvd.py index a1bc085..0af3c06 100644 --- a/polara/lib/hosvd.py +++ b/polara/lib/hosvd.py @@ -27,16 +27,16 @@ def tensordot2(idx, val, shape, U, V, modes): ten_mode1, mat_mode1 = modes[0] ten_mode2, mat_mode2 = modes[1] - ten_mode0, = [x for x in (0,1,2) if x not in (ten_mode1, ten_mode2)] + ten_mode0, = [x for x in (0, 1, 2) if x not in (ten_mode1, ten_mode2)] new_shape = (shape[ten_mode0], U.shape[1-mat_mode1], V.shape[1-mat_mode2]) res = np.zeros(new_shape) - if mat_mode1==1: + if mat_mode1 == 1: vU = U.T else: vU = U - if mat_mode2==1: + if mat_mode2 == 1: vV = V.T else: vV = V diff --git a/polara/recommender/coldstart/data.py b/polara/recommender/coldstart/data.py index 9038e12..b29efcf 100644 --- a/polara/recommender/coldstart/data.py +++ b/polara/recommender/coldstart/data.py @@ -97,7 +97,7 @@ def _reindex_cold_items(self): item_index = self.index.itemid new_item_index = (namedtuple('ItemIndex', 'training cold_start') - ._make([item_index, cold_item_index])) + ._make([item_index, cold_item_index])) self.index = self.index._replace(itemid=new_item_index) diff --git a/polara/recommender/data.py b/polara/recommender/data.py index 69ec83a..ae06c55 100644 --- a/polara/recommender/data.py +++ b/polara/recommender/data.py @@ -217,7 +217,7 @@ def prepare_training_only(self): def _validate_config(self): if self._warm_start and not (self._holdout_size and self._test_ratio): raise ValueError('Both holdout_size and test_ratio must be positive when warm_start is set to True') - if not self._warm_start and (self._holdout_size==0) and (self._test_ratio>0): + if not self._warm_start and (self._holdout_size == 0) and (self._test_ratio > 0): raise ValueError('test_ratio cannot be nonzero when holdout_size is 0 and warm_start is set to False') assert self._test_ratio < 1, 'Value of test_ratio can\'t be greater than or equal to 1' @@ -512,7 +512,7 @@ def _filter_short_sessions(self): holdout_sessions = holdout.groupby(userid, sort=False) holdout_sessions_len = holdout_sessions.size() - invalid_sessions = (holdout_sessions_len!=self.holdout_size) + invalid_sessions = (holdout_sessions_len != self.holdout_size) if invalid_sessions.any(): n_invalid_sessions = invalid_sessions.sum() invalid_session_index = invalid_sessions.index[invalid_sessions] diff --git a/polara/recommender/external/mymedialite/mmlwrapper.py b/polara/recommender/external/mymedialite/mmlwrapper.py index 1b0d4b2..0bec58e 100644 --- a/polara/recommender/external/mymedialite/mmlwrapper.py +++ b/polara/recommender/external/mymedialite/mmlwrapper.py @@ -122,8 +122,8 @@ def _remap_factors(entity_mapping, entity_factors, num_entities, num_factors): def _parse_factors(self): model_data_path = self.saved_model_path - model_params = pd.read_csv(model_data_path, skiprows=2, sep=' ', - header=None, names=['col1', 'col2', 'col3']) + model_params = pd.read_csv(model_data_path, skiprows=2, header=None, + sep=' ', names=['col1', 'col2', 'col3']) num_users = self.data.index.userid.training.new.max() + 1 num_items = self.data.index.itemid.new.max() + 1 @@ -143,8 +143,8 @@ def _parse_factors(self): NotImplementedError('{} data is not recognized.'.format(model_data_path)) if self.positive_only: - user_mapping = pd.read_csv(self.user_mapping_file, sep = '\t', header=None) - item_mapping = pd.read_csv(self.item_mapping_file, sep = '\t', header=None) + user_mapping = pd.read_csv(self.user_mapping_file, sep='\t', header=None) + item_mapping = pd.read_csv(self.item_mapping_file, sep='\t', header=None) user_factors_full = self._remap_factors(user_mapping, users_factors, num_users, nf) item_factors_full = self._remap_factors(item_mapping, items_factors, num_items, nf) diff --git a/polara/recommender/models.py b/polara/recommender/models.py index cb35e88..e05a797 100644 --- a/polara/recommender/models.py +++ b/polara/recommender/models.py @@ -251,8 +251,8 @@ def _make_user(self, user_info): # conversion is not required for feedback (it's made in *to_coo functions, if needed) items_data = item_index.set_index('old').loc[items_data, 'new'].values user_data = pd.DataFrame({userid: [0]*len(items_data), - itemid: items_data, - feedback: feedback_data}) + itemid: items_data, + feedback: feedback_data}) return user_data From 95a29adb764cd586b0ed1a3cc78241eed99014c2 Mon Sep 17 00:00:00 2001 From: Evgeny Frolov Date: Sun, 11 Mar 2018 14:15:34 +0300 Subject: [PATCH 11/21] fix python 2 dict iteration --- polara/recommender/data.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/polara/recommender/data.py b/polara/recommender/data.py index ae06c55..b2ea411 100644 --- a/polara/recommender/data.py +++ b/polara/recommender/data.py @@ -44,7 +44,7 @@ def unsubscribe(self, event, subscriber): del self._get_subscribers(event)[subscriber] def unsubscribe_any(self, subscriber): - for event in self._subscribers.iterkeys(): + for event in self._subscribers: subscribers = self._get_subscribers(event) if subscriber in subscribers: del subscribers[subscriber] From 7fcf22dd9802678d864a8d4e04464e703591f08c Mon Sep 17 00:00:00 2001 From: Evgeny Frolov Date: Sun, 11 Mar 2018 14:15:56 +0300 Subject: [PATCH 12/21] fix unsafe default argument --- polara/recommender/data.py | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/polara/recommender/data.py b/polara/recommender/data.py index b2ea411..53e7fb6 100644 --- a/polara/recommender/data.py +++ b/polara/recommender/data.py @@ -20,11 +20,12 @@ def random_sample(df, frac, random_state): class EventNotifier(object): - def __init__(self, events=[]): + def __init__(self, events=None): self._subscribers = {} - assert isinstance(events, list) - for event in events: - self.register_event(event) + if events is not None: + assert isinstance(events, list) + for event in events: + self.register_event(event) def register_event(self, event): self._subscribers[event] = WeakKeyDictionary({}) From e16689ea23065d0cc08c56991e00ad7dd02b2c42 Mon Sep 17 00:00:00 2001 From: Evgeny Frolov Date: Sun, 11 Mar 2018 14:16:40 +0300 Subject: [PATCH 13/21] fix class naming convention in metaclass construction --- polara/recommender/models.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/polara/recommender/models.py b/polara/recommender/models.py index e05a797..ff0f3de 100644 --- a/polara/recommender/models.py +++ b/polara/recommender/models.py @@ -58,8 +58,8 @@ class MetaModel(type): # performs cleaning of the instance when build method is called # propagates the action to any subclasses, key idea is borrowed from here: # https://stackoverflow.com/questions/18858759/python-decorating-a-class-method-that-is-intended-to-be-overwritten-when-inheri - def __new__(meta, name, bases, clsdict): - cls = super(MetaModel, meta).__new__(meta, name, bases, clsdict) + def __new__(mcs, name, bases, clsdict): + cls = super(MetaModel, mcs).__new__(mcs, name, bases, clsdict) if 'build' in clsdict: setattr(cls, 'build', clean_build_decorator(clsdict['build'])) return cls From 0927d89a82d572e17adc4a8aa2e7a48ed96d46e1 Mon Sep 17 00:00:00 2001 From: Evgeny Frolov Date: Sun, 11 Mar 2018 14:17:19 +0300 Subject: [PATCH 14/21] make non-instance-dependent method a static method --- polara/recommender/models.py | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/polara/recommender/models.py b/polara/recommender/models.py index ff0f3de..d618c02 100644 --- a/polara/recommender/models.py +++ b/polara/recommender/models.py @@ -201,7 +201,8 @@ def _get_test_data(self): return test_data, test_shape, test_users - def _slice_test_data(self, test_data, start, stop): + @staticmethod + def _slice_test_data(test_data, start, stop): user_coo, item_coo, fdbk_coo = test_data slicer = (user_coo >= start) & (user_coo < stop) From 8e89771cc87fc3a8befc07d568acb804f7d9dc7c Mon Sep 17 00:00:00 2001 From: Evgeny Frolov Date: Sun, 11 Mar 2018 14:18:03 +0300 Subject: [PATCH 15/21] fix incosistent arguments in slice_recommendation method of a subclass --- polara/recommender/models.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/polara/recommender/models.py b/polara/recommender/models.py index d618c02..7d873b3 100644 --- a/polara/recommender/models.py +++ b/polara/recommender/models.py @@ -621,7 +621,7 @@ def _sparse_dot(self, tst_mat, i2i_mat): return scores - def slice_recommendations(self, test_data, shape, start, stop, test_user=None): + def slice_recommendations(self, test_data, shape, start, stop, test_users=None): test_matrix, slice_data = self.get_test_matrix(test_data, shape, (start, stop)) # NOTE CSR format is mandatory for proper handling of signle user # recommendations, as vector of shape (1, N) in CSC format is inefficient From 7f94e30512109fb25e3f8cd855a624a1dea4143f Mon Sep 17 00:00:00 2001 From: Evgeny Frolov Date: Sun, 11 Mar 2018 14:18:20 +0300 Subject: [PATCH 16/21] simplify isinstance check --- polara/recommender/models.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/polara/recommender/models.py b/polara/recommender/models.py index 7d873b3..c0b4fc8 100644 --- a/polara/recommender/models.py +++ b/polara/recommender/models.py @@ -730,7 +730,7 @@ def flatten_scores(tensor_scores, flattener=None): elif isinstance(flattener, int): slicer = flattener matrix_scores = tensor_scores[:, :, slicer] - elif isinstance(flattener, list) or isinstance(flattener, slice): + elif isinstance(flattener, (list, slice)): slicer = flattener flatten = np.sum matrix_scores = flatten(tensor_scores[:, :, slicer], axis=-1) From ba1c9bd3bc306665168963eb2236e1a10dd2a494 Mon Sep 17 00:00:00 2001 From: Evgeny Frolov Date: Sun, 11 Mar 2018 14:19:11 +0300 Subject: [PATCH 17/21] align imports with importing order convention --- polara/recommender/external/mymedialite/mmlwrapper.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/polara/recommender/external/mymedialite/mmlwrapper.py b/polara/recommender/external/mymedialite/mmlwrapper.py index 0bec58e..cf4cf7c 100644 --- a/polara/recommender/external/mymedialite/mmlwrapper.py +++ b/polara/recommender/external/mymedialite/mmlwrapper.py @@ -1,9 +1,9 @@ -from polara.recommender.models import SVDModel from subprocess import call from shlex import split import sys import pandas as pd import numpy as np +from polara.recommender.models import SVDModel #default settings From abe5a19288a2bf3cb8ea2772591040da3dd2ba7e Mon Sep 17 00:00:00 2001 From: Evgeny Frolov Date: Sun, 11 Mar 2018 14:21:56 +0300 Subject: [PATCH 18/21] update and fix data preprocessing routines --- polara/datasets/bookcrossing.py | 9 ++++++--- polara/datasets/movielens.py | 21 ++++++++++++++------- 2 files changed, 20 insertions(+), 10 deletions(-) diff --git a/polara/datasets/bookcrossing.py b/polara/datasets/bookcrossing.py index a2ef145..8a24bba 100644 --- a/polara/datasets/bookcrossing.py +++ b/polara/datasets/bookcrossing.py @@ -26,21 +26,24 @@ def get_bx_data(local_file=None, get_ratings=True, get_users=False, get_books=Fa delimiter = ';' if get_ratings: zdata = zfile.read(zip_file) - ratings = pd.read_csv(BytesIO(zdata), sep=delimiter, header=0, engine='c') + ratings = pd.read_csv(BytesIO(zdata), sep=delimiter, header=0, + engine='c', encoding='unicode_escape') if get_users: zip_file = zip_files[zip_files.str.contains('users', flags=2)].iat[0] with zfile.open(zip_file) as zdata: - users = pd.read_csv(zdata, sep=delimiter, header=0, engine='c',) + users = pd.read_csv(zdata, sep=delimiter, header=0, engine='c', + encoding='unicode_escape') if get_books: zip_file = zip_files[zip_files.str.contains('books', flags=2)].iat[0] with zfile.open(zip_file) as zdata: books = pd.read_csv(zdata, sep=delimiter, header=0, engine='c', - quoting=1, escapechar='\\', + quoting=1, escapechar='\\', encoding='unicode_escape', usecols=['ISBN', 'Book-Author', 'Publisher']) res = [data.rename(columns=lambda x: x.lower().replace('book-', '') .replace('-id', 'id'), copy=False) for data in [ratings, users, books] if data is not None] + if len(res)==1: res = res[0] return res diff --git a/polara/datasets/movielens.py b/polara/datasets/movielens.py index 3d8d84e..aae4549 100644 --- a/polara/datasets/movielens.py +++ b/polara/datasets/movielens.py @@ -26,21 +26,28 @@ def get_movielens_data(local_file=None, get_ratings=True, get_genres=False, zip_files = pd.Series(zfile.namelist()) zip_file = zip_files[zip_files.str.contains('ratings')].iat[0] is_latest = 'latest' in zip_file - header = 0 if is_latest else None + is_20m = '20m' in zip_file + delimiter = ',' + header = 0 if (is_latest or is_20m) else None if get_ratings: zdata = zfile.read(zip_file) - delimiter = ',' - zdata = zdata.replace(b'::', delimiter.encode()) # makes data compatible with pandas c-engine + zdata = zdata.replace(b'::', delimiter.encode()) + # makes data compatible with pandas c-engine + # returns string objects instead of bytes in that case ml_data = pd.read_csv(BytesIO(zdata), sep=delimiter, header=header, engine='c', names=['userid', 'movieid', 'rating', 'timestamp'], usecols=['userid', 'movieid', 'rating']) if get_genres: zip_file = zip_files[zip_files.str.contains('movies')].iat[0] - with zfile.open(zip_file) as zdata: - delimiter = ',' if is_latest else '::' - genres_data = pd.read_csv(zdata, sep=delimiter, header=header, engine='python', - names=['movieid', 'movienm', 'genres']) + zdata = zfile.read(zip_file) + if not is_latest: + # make data compatible with pandas c-engine + # pandas returns string objects instead of bytes in that case + zdata = zdata.replace(b'::', delimiter.encode()) + genres_data = pd.read_csv(BytesIO(zdata), sep=delimiter, header=header, + engine='c', encoding='unicode_escape', + names=['movieid', 'movienm', 'genres']) ml_genres = get_split_genres(genres_data) if split_genres else genres_data From 4fee75b97a89a635a6f60ba45069e315561524fb Mon Sep 17 00:00:00 2001 From: Evgeny Frolov Date: Sun, 11 Mar 2018 14:35:41 +0300 Subject: [PATCH 19/21] correct shape of tensor unfolding --- polara/recommender/models.py | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/polara/recommender/models.py b/polara/recommender/models.py index c0b4fc8..7b46965 100644 --- a/polara/recommender/models.py +++ b/polara/recommender/models.py @@ -786,10 +786,9 @@ def slice_recommendations(self, test_data, shape, start, stop, test_users=None): num_users = stop - start num_items = shape[1] - num_fdbks = shape[2] # assume that w.shape[1] < v.shape[1] (allows for more efficient calculations) - scores = test_tensor_unfolded.dot(w).reshape(num_users, num_items, num_fdbks) + scores = test_tensor_unfolded.dot(w).reshape(num_users, num_items, w.shape[1]) scores = np.tensordot(scores, v, axes=(1, 0)) scores = np.tensordot(np.tensordot(scores, v, axes=(2, 1)), w, axes=(1, 1)) scores = self.flatten_scores(scores, self.flattener) From c572e8da5f06bac1878fb7b6429d4fba8135ca99 Mon Sep 17 00:00:00 2001 From: Evgeny Frolov Date: Sun, 11 Mar 2018 14:42:28 +0300 Subject: [PATCH 20/21] update examples --- examples/Example_ML1M.ipynb | 240 ++++++++++++++++++------------------ 1 file changed, 120 insertions(+), 120 deletions(-) diff --git a/examples/Example_ML1M.ipynb b/examples/Example_ML1M.ipynb index 6df36af..bd79d5f 100644 --- a/examples/Example_ML1M.ipynb +++ b/examples/Example_ML1M.ipynb @@ -23,7 +23,7 @@ }, { "cell_type": "code", - "execution_count": 1, + "execution_count": 2, "metadata": { "ExecuteTime": { "end_time": "2017-01-19T00:18:07.398000", @@ -44,7 +44,7 @@ "%matplotlib inline\n", "\n", "from polara.recommender.data import RecommenderData\n", - "from polara.recommender.models import SVDModel, CoffeeModel, NonPersonalized\n", + "from polara.recommender.models import SVDModel, CoffeeModel, PopularityModel, RandomModel\n", "from polara.evaluation import evaluation_engine as ee\n", "from polara.recommender.external.mymedialite.mmlwrapper import MyMediaLiteWrapper\n", "from polara.datasets.movielens import get_movielens_data\n", @@ -75,7 +75,7 @@ }, { "cell_type": "code", - "execution_count": 2, + "execution_count": 3, "metadata": { "ExecuteTime": { "end_time": "2017-01-19T00:18:09.900000", @@ -99,7 +99,7 @@ }, { "cell_type": "code", - "execution_count": 3, + "execution_count": 4, "metadata": { "ExecuteTime": { "end_time": "2017-01-19T00:18:10.726000", @@ -127,7 +127,7 @@ }, { "cell_type": "code", - "execution_count": 4, + "execution_count": 7, "metadata": { "ExecuteTime": { "end_time": "2017-01-19T00:18:14.527000", @@ -148,7 +148,7 @@ }, { "cell_type": "code", - "execution_count": 7, + "execution_count": 8, "metadata": { "ExecuteTime": { "end_time": "2017-01-19T00:18:16.284000", @@ -165,7 +165,7 @@ }, { "cell_type": "code", - "execution_count": 8, + "execution_count": 9, "metadata": {}, "outputs": [ { @@ -182,7 +182,7 @@ " 'warm_start': True}" ] }, - "execution_count": 8, + "execution_count": 9, "metadata": {}, "output_type": "execute_result" } @@ -200,7 +200,7 @@ }, { "cell_type": "code", - "execution_count": 9, + "execution_count": 10, "metadata": { "ExecuteTime": { "end_time": "2017-01-19T00:18:20.954000", @@ -212,14 +212,14 @@ "# bpr = MyMediaLiteWrapper(LIB_PATH, MML_DATA, 'BPRMF', data_model)\n", "# wrmf = MyMediaLiteWrapper(LIB_PATH, MML_DATA, 'WRMF', data_model)\n", "svd = SVDModel(data_model)\n", - "popular = NonPersonalized('mostpopular', data_model)\n", - "random = NonPersonalized('random', data_model)\n", + "popular = PopularityModel(data_model)\n", + "random = RandomModel(data_model, seed=0)\n", "coffee = CoffeeModel(data_model)" ] }, { "cell_type": "code", - "execution_count": 10, + "execution_count": 11, "metadata": { "ExecuteTime": { "end_time": "2017-01-19T00:19:37.710000", @@ -233,7 +233,7 @@ }, { "cell_type": "code", - "execution_count": 11, + "execution_count": 12, "metadata": { "ExecuteTime": { "end_time": "2017-01-19T00:19:40.427000", @@ -244,10 +244,10 @@ { "data": { "text/plain": [ - "['PureSVD', 'CoFFee', 'mostpopular', 'random']" + "['PureSVD', 'CoFFee', 'MP', 'RND']" ] }, - "execution_count": 11, + "execution_count": 12, "metadata": {}, "output_type": "execute_result" } @@ -261,7 +261,7 @@ }, { "cell_type": "code", - "execution_count": 12, + "execution_count": 13, "metadata": { "ExecuteTime": { "end_time": "2017-01-19T00:19:41.991000", @@ -276,7 +276,7 @@ }, { "cell_type": "code", - "execution_count": 13, + "execution_count": 14, "metadata": { "ExecuteTime": { "end_time": "2017-01-19T00:19:42.483000", @@ -290,8 +290,8 @@ "text": [ "PureSVD 4\n", "CoFFee 4\n", - "mostpopular 4\n", - "random 4\n" + "MP 4\n", + "RND 4\n" ] } ], @@ -302,7 +302,7 @@ }, { "cell_type": "code", - "execution_count": 14, + "execution_count": 15, "metadata": { "ExecuteTime": { "end_time": "2017-01-19T00:19:43.197000", @@ -340,7 +340,7 @@ }, { "cell_type": "code", - "execution_count": 15, + "execution_count": 16, "metadata": { "ExecuteTime": { "end_time": "2017-01-19T00:19:46.148000", @@ -356,7 +356,7 @@ }, { "cell_type": "code", - "execution_count": 16, + "execution_count": 17, "metadata": { "ExecuteTime": { "end_time": "2017-01-19T00:19:51.117000", @@ -378,7 +378,7 @@ }, { "cell_type": "code", - "execution_count": 19, + "execution_count": 18, "metadata": { "ExecuteTime": { "end_time": "2017-01-19T00:19:57.864000", @@ -403,8 +403,8 @@ "1 of 1208 userid's were filtered out from holdout. Reason: not enough items.\n", "1 userid's were filtered out from testset. Reason: inconsistent with holdout.\n", "Done.\n", - "PureSVD training time: 0.10939731338624485s\n", - "CoFFee training time: 1.9994920114886359s\n", + "PureSVD training time: 0.10657037373075795s\n", + "CoFFee training time: 1.5177289084482908s\n", "100 70 50 30 20 15 10 5 3 2 1 \n", "============ Fold: 2 =============\n", "Preparing data...\n", @@ -413,8 +413,8 @@ "2 of 1208 userid's were filtered out from holdout. Reason: not enough items.\n", "2 userid's were filtered out from testset. Reason: inconsistent with holdout.\n", "Done.\n", - "PureSVD training time: 0.09770373147282374s\n", - "CoFFee training time: 1.6519600406682544s\n", + "PureSVD training time: 0.12100629132648777s\n", + "CoFFee training time: 2.135373650530159s\n", "100 70 50 30 20 15 10 5 3 2 1 \n", "============ Fold: 3 =============\n", "Preparing data...\n", @@ -423,8 +423,8 @@ "4 of 1208 userid's were filtered out from holdout. Reason: not enough items.\n", "4 userid's were filtered out from testset. Reason: inconsistent with holdout.\n", "Done.\n", - "PureSVD training time: 0.09997345185669104s\n", - "CoFFee training time: 2.7146723711138705s\n", + "PureSVD training time: 0.0964248257231084s\n", + "CoFFee training time: 1.3818089788154566s\n", "100 70 50 30 20 15 10 5 3 2 1 \n", "============ Fold: 4 =============\n", "Preparing data...\n", @@ -433,8 +433,8 @@ "2 of 1208 userid's were filtered out from holdout. Reason: not enough items.\n", "2 userid's were filtered out from testset. Reason: inconsistent with holdout.\n", "Done.\n", - "PureSVD training time: 0.09677126440637807s\n", - "CoFFee training time: 1.7942675702759985s\n", + "PureSVD training time: 0.09338965818536238s\n", + "CoFFee training time: 2.131201799438987s\n", "100 70 50 30 20 15 10 5 3 2 1 \n", "============ Fold: 5 =============\n", "Preparing data...\n", @@ -443,8 +443,8 @@ "2 of 1208 userid's were filtered out from holdout. Reason: not enough items.\n", "2 userid's were filtered out from testset. Reason: inconsistent with holdout.\n", "Done.\n", - "PureSVD training time: 0.09881926403295438s\n", - "CoFFee training time: 2.2371861447865733s\n", + "PureSVD training time: 0.11019193432397856s\n", + "CoFFee training time: 1.4080982047912087s\n", "100 70 50 30 20 15 10 5 3 2 1 " ] } @@ -464,7 +464,7 @@ }, { "cell_type": "code", - "execution_count": 20, + "execution_count": 19, "metadata": { "ExecuteTime": { "end_time": "2017-01-19T00:19:57.868000", @@ -478,7 +478,7 @@ }, { "cell_type": "code", - "execution_count": 21, + "execution_count": 20, "metadata": {}, "outputs": [], "source": [ @@ -499,7 +499,7 @@ }, { "cell_type": "code", - "execution_count": 22, + "execution_count": 21, "metadata": { "ExecuteTime": { "end_time": "2017-01-19T00:18:13.045000", @@ -520,12 +520,12 @@ }, { "cell_type": "code", - "execution_count": 23, + "execution_count": 22, "metadata": {}, "outputs": [ { "data": { - "image/png": "iVBORw0KGgoAAAANSUhEUgAAAlkAAAFACAYAAACPyWmJAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4wLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvFvnyVgAAIABJREFUeJzs3Xd4lFXax/HvmT7JpAEBhNBrIBCBANJkQaQpooAs4tp2XfvqWlbXta2uva3dXXR9d91qb4SOFOlEkBpC7x3SM33O+8dEZDKTZCjJJHB/rstLM/M8zxwkmfnlnPu5j9JaI4QQQgghzi5DrAcghBBCCHEukpAlhBBCCFEDJGQJIYQQQtQACVlCCCGEEDVAQpYQQgghRA2QkCWEEEIIUQMkZAkhhBBC1AAJWUIIIYQQNUBClhBCCCFEDTDFegCnqlGjRrp169axHoYQIsb8WhMIgNmoYj0UIc46v99PUVERTqeTAwcOHNVap8Z6TOLU1buQ1bp1a3JycmI9DCFEjLi8fopdPgLlW4KlOqwYDBK0xLnB7/ezbNkyFixYgNaaiy++mIsvvnhXrMclTk+9C1lCiPOTP6Apdnlx+wKxHooQNebjjz9m8+bNdOrUiZEjR5KcnBzrIYkzICFLCFHnlbp9lLp9VNzO3h/QHC5y0TTZHpNxCXE2lJSUYLVaMZvN9O3bl549e9KpU6dYD0ucBRKyhBB1ltcfoMjpxReoGK9gx9FSnsnOxesP8M1vBmI2yn08on4JBALk5OTw7bff0qdPH4YOHUrbtm1jPSxxFknIEkLUOVprit0+nB5/2HM+f4B/LtvF3xbtwG4x8sfLu2KSmixRz+zbt4/s7GwOHDhAmzZt6N69e6yHJGqAhCwhRJ1SsbD9ZLkHing6O5eth0sYlt6Y+4d3olOTBJSSkCXqj2XLljFz5kwcDgfjx4+na9eu8j18jpKQJYSoE6oqbHd5/fxt0Q7+vWw3KfFmXhzfncGd5I52UX9orfF6vVgsFtq0aUPfvn0ZMmQIVqs11kMTNUhClhAi5iorbAdYvTufZ6blsue4k7EXNuM3Q9uTYDPX+hiFOF2HDx8mOzubpKQkxo0bR5MmTRg5cmSshyVqgYQsIUTMeHwBil2RC9tL3D7embeVz1bto1myjbeu6UHvNg1iMEohTo/b7WbBggUsW7YMm81GZmZmrIckapmELCHOR1qDqxDssenBU1VhO8DirUd5YcYmDhe5uaZPC269uB12izHkGAU4bCZpRCrqpL179/Lxxx9TXFxMz549ueSSS4iLi4v1sEQtk5AlxPnG7wNnPgR8MQlZVRW2F5R5+POcLcxYf5A2jeJ574ZudGueFHac2Wgg0WbCJG0bRB2jtUYpRXJyMg0aNGDixImkpaXFelgiRiRkCXE+8brAVRCcyaplVRW2a62Zm3uYl2flUeTy8auBbbixf2ssptAQpYB4q4l4q7x1ibrF6/WyaNEi9uzZw3XXXYfD4eDGG2+M9bBEjMk7lRDnC1cReEpj8tJVFbYfKXbz4sxNLNx8lPQLEnhzcjodGieEHSezV6Ku2rJlC9OnTyc/P5+MjIwTdxEKISFLiHNdwA/OAvB7av2lqyps11rzzZoDvD53C15/gN8Mbc+kPi0wGWT2StQPZWVlTJ06ldzcXBo2bMj1119PmzZtYj0sUYfIu5YQ5zKfJ1h/pWt3U+XqCtv35Tt5dlouObvy6dkymT+MTqdFg/CiYJNBkWQ3y+yVqJPMZjNHjx5l6NCh9O/fH6PRWP1J4rwiIUuIc5W7BNzFtf6yVRW2+wOaj3P28O78bRgNit+P6szYC5thqNDtWgFxVhMOmb0SdcyuXbtYsmQJEyZMwGw2c9ttt2EwyC8BIjJ5BxPiXKN1cPbK567Vl62qsB1g+5ESns7OZcP+Iga0b8hDIzvTJNEWdpzJoEi0m2XDZ1GnlJSUMHv2bNauXUtSUhIFBQWkpqZKwBJVkpAlxLnE7w3WXwV81R/rKgr+O/GCM37Zqgrbvf4A/1iyk/9bvBOH1cRTY7syvEuTiHu1xVtNxFuMso+bqDO01uTk5DB37ly8Xi+DBg1i0KBBmM2y64ConoQsIc4VXmewwWg07RmObIKv74ak5nDTdDjNUFNVYTvAxv1FPJOdy9YjJQzv0oT7Lu1ISnz4XVfG8tormb0SddHatWtp1qwZo0ePplGjRrEejqhHJGQJUd9pDe4i8JRFd/yGz2HOk2BLgmF/PK2AVV1hu8vrZ8rC7fx3xW4aOqy8fHV3BnWIvKFznMWIw2qS2StRZzidThYsWMDAgQNxOBxMnjwZm80m36PilEnIEqI+C/iD9Vd+b/XH+tww72lY9wm06AuXvQpNM075JasqbAf4flc+z07LZW++k6t6NOeuIe1x2MLfaowGRaLNHNZwVIhY0Vqzdu1aZs2ahdPppHnz5nTr1g273R7roYl6SkKWEPWVzx2sv4qmPUPhXvjmbji8EfrcAv3vBsOp/fhXV9he4vLx5rdb+PKH/aSl2Hl7cg+yWkfe0NluMZIgs1eiDjl8+DDZ2dns3r2btLQ0LrvsMpo2bRrrYYl6TkKWEPWRuzjYoiEa2+fD9AcBDWPfgXZDT/nlqipsB1i05SjPz9jEsRI31/ZtyS0Xt8VmDu8ZZFDB2iuZvRJ1zaJFizhy5AhjxoyhR48e8guAOCskZAlRnwQCwb0Ho2nPEPDDkjdgxV+hcRe4/HVIbnFKL1ddYXt+qYdXZ29m1sZDtEuN54Xx3ejaLHxDZ5DZK1G3aK3Jzc2lUaNGNG7cmBEjRjBy5Eji4sKb4gpxuiRkCVFf+L3B+qtA5GLzEGXHYNoDsHspZEyAIY+CObwnVWUCAU2Jp/LCdq01szYe4tVZmylx+/j1oDbc0L91xLsDDUqRaDdhNUk3bFE3HD9+nOnTp7N161Z69uzJmDFjiI+Pj/WwxDlIQpYQ9YGnNLhEGE17hv2rYOq9wXqt4c9AxvhTeqnqCtsPFbl4cUYei7YepWuzRB4ZnU67xo6Ix9rMRhJtMnsl6gafz8eiRYtYtGgRRqORkSNH0rt371gPS5zDJGQJUZdpHVwe9LqiO3b1P2Hhi5BwAVzz3+AyYZSqK2wPaM1XP+znzW+34PNrfjusAxOzWmA0hAcomb0SddHy5ctZsGABGRkZDB8+nISEhFgPSZzjJGQJUVf5feXLg1F0b/eUwKzHYPP0YGH7iOfBlhj1S1VX2L7neBnPTstl1e4Cslql8IfR6TRPiXxbu80crL0yRAhfQtS2oqIiSkpKaNasGX369KFZs2a0adMm1sMS5wkJWULURV5XcAYrmuXBY1uD7Rnyd8LA+6H3r0BFcfeeyVptYbsvEOB/K/YwZeF2TEbFH0Z35orMZhGX/5SCRJs54l2FQtQ2v9/P8uXLmT9/PikpKdx2222YzWYJWKJWScgSoq5xFQVrsKKxaSrMfhzMdhj/AbS8KKrTtNVBsbbhLPNUeszWwyU8k53LxgNFDOrQiAdHdqJxQuTieZvJSIJNZq9E3bBr1y6ys7M5cuQIHTt2ZOTIkVIXKGJCQpYQdUUgUN69vfLgc4LfAwtegB/+Dc16wuV/BkeT6s9TBvzWJAq9Brz+yHcOenwB/r5kJ39fspNEm4mnr8xgWHpjmb0S9cKOHTv48MMPSUpKYtKkSXTq1CnWQxLnsRoNWUqpkcDrgBF4X2v9fIXnbwReAvaVP/SW1vr9mhyTEHWSzxMMWNF0by8+AFN/CwfWQM8bYNADYDRXf57RgsecSIHbj67kddbvK+SZ7Fy2Hy1lZEZT7h3WgeS48A2dAawmA4k2s8xeiZgLBAIcO3aM1NRUWrduzahRo7jwwguxWCJ/7wpRW2osZCmljMDbwKXAXmClUuprrfXGCod+pLW+q6bGIUSd5y4JtmeIxq7Fwf5Xfg9c/hp0HBndeZZ4SlUcJa7IRfROj5+/LtzG/1bsITXByqsTMxnQvlHEY2X2StQl+/fvJzs7m+PHj/Ob3/yGuLg4+vTpE+thCQHU7ExWH2Cr1no7gFLqf8BYoGLIEuL8pHVw9iqa7u06AMv/AkvehIbtYczr0KBt9ecphbYlUeQ14fJFDlg5O4/z7LRN7CtwMr5nc+4Y0h6HNfJbg8xeibrC6XTy7bffkpOTg8PhYPTo0bKRs6hzajJkNQf2nPT1XqBvhOPGK6UuBjYD92qt91Q8QCl1C3ALQMuWLWtgqELUMr832Cw0mvYMzoLg3oM7F0LnMXDpk2COYusPoxmfJZECt8YfoUt8scvLm99u5asf9tOigZ2//KInPVqmRLyUUpBgNWO3yOyViL3S0lLeffddysrK6NOnD0OGDMFmi35HAyFqS02GrEi/6la8T/wb4L9aa7dS6jbgH0DY7rVa6ynAFICsrKwo7mkXog7zOsFVGF17hoPrYOo9UHIELnkCuk8KJp7qmG24jAkUuXwRX+b7Xfk88dUGjpW6ue6iVtw8qE2ly38Wo4FEuzli01EhalNZWRlxcXHEx8fTu3dvOnbsyAUXXBDrYQlRqZoMWXuBk3ejTQP2n3yA1vrYSV++B7xQg+MRIra0BncReMqiO3bdRzDvGYhLhZ//Gy7oXv15SoE1gRJtpTRC/ZXWmn8t280787fSIiWOl67uTfoFkZuWKiDBJrNXIvY8Hg8LFixgxYoV/PrXv6Zx48YMHjw41sMSolo1GbJWAh2UUm0I3j04CZh88gFKqQu01gfKv7wCyK3B8QgROwF/eXsGb/XHep0w94+w8StoNRBGvwT2yMt4IQxGAtZkCj3g8YcHrBKXj6embmTB5iNc0rkxj1yWTnwltVcyeyXqAq01mzZtYsaMGRQVFdGjRw8cjsj7ZApRF9VYyNJa+5RSdwEzCbZw+EBrvUEp9RSQo7X+GrhbKXUF4AOOAzfW1HiEiBmfO1hXFU17hvydwe7tR7dAv7ug7+1giGImyWTFa0mkwOmPuLHzlsPF/P6zdRwodPHbYR2Y1LtF5L5XgMNmIs4iLfREbGmt+eijj8jLy6NJkyZMmDCBFi1aVH+iEHWI0tHUhdQhWVlZOicnJ9bDECI67uJgi4ZobJkNMx8OhqrRL0PrQdGdZ3XgVHEUu7wR9x7MXnuAF2ZsItFm5pmrMshskRzxMmajgUSbCZMxii15hKghfr8fozH4i8XChQuxWCz06dMHg+H8/b5USn2vtc6K9TjEqZNfV4WoCYFAcO/BaNozBHzw3avw/QfQpBuMeQ0Sm1d/njKgbUkU+4043eHLkG6fn1dnbebLH/bTq1UKfxrblYYOa/hlkNkrUTds3bqV6dOnM3r0aNq1a8fFF18c6yEJcUbkXVWIs83vA+fxYB1WdUoOQ/Z9sC8HMifD4N+DKYou1UZLcHscdyDi9jj7C5w8/Pk6Nh0s5vp+rbh1cFtMEWYCZPZK1AVFRUXMnDmTjRs30qBBA0wm+WgS5wb5ThbibPK6gjNY0SzD71kRDFieUhj1EqSPie41LHG4jfEUOiO3Z1iy7ShPfLWBgIYXJ3RncMfUiJdxWE2VFr4LUVtycnKYNWsWWmuGDBlC//79JWSJc4Z8JwtxtkRbf6U15HwAi16F5BYw4QNo1LH685QCWxKlATMlzvC7B/0Bzd8W7eCDRTto19jB8+O60aJBeNNSg1Ik2c1YTDJ7JeqGH/cbTEmJ4i5aIeoRCVlCnKlT2R7HXQwz/wBbZ0OHETD8GbBGcUu6wYS2J1PkJuL2OAVlHh7/agPLdxznsu4X8OCIThGbi1qMBpLssi2OiJ3S0lLmzJlDWloavXr1OvFPpLtdhajvJGQJcSZOpf/Vkbxge4bCvcHaq543RN293WdOpMDpwx8IXx/csL+Qhz9fR36plz+M7swVmc0ifmDFW02V7kkoRE0LBAKsWrWKuXPn4vF4aNiwIYCEK3FOk3dcIU7XqfS/2vAFzH0SrAkw8UNo3qv6c8q7t7uUjaKy8PYMWms+W7WPP8/eTGqClSnX94rYvV0pSLKbsZqkc7uIjYMHD/LNN9+wf/9+WrduzejRo0lNjVwrKMS5REKWEKfDUwquouqP87mDW+Os+xha9IXRr0B8o+rPMxjBlkyxT1HmCZ8lc3r8PD99EzM2HKR/u4b88YquJNnNYceZy5cHpXO7iKXS0lIKCwsZN24cGRkZMnslzhvSjFSIU6F1cHNnr7P6Ywv3Bjd3PrQBev8aBtwDhih+rzFZCViTKHT58fjDZ8l2Hyvjoc/WsuNoKbdc3JYbB7TGEOFDK85iJMEWHryEqGlaa9atW0dxcTEDBgwAwOv1YjbL9+PpkGak9ZfMZAkRrUCgvP7KU/2x2xfA9AeBAIx9B9oNje41rA68pngKyrwRt8f5dtNh/jR1I2ajgdevuZC+bRqGHaMUJNrMEQvfhahpR44cITs7m127dtGyZUv69euHwWCQgCXOSxKyhIiG3xsMWNU1GNUBWPo2LHsHUjvBmDcguWX111cGsCfjDJgoLvWE1V/5/AHenr+N/yzfTddmiTw3rhtNEm1hlzEZFMlxFlkeFLXO4/GwcOFCli5disVi4fLLL6dnz56yNCjOaxKyhKiO1xlcIqxuad1VGJy92rEAuoyFS/4IZnv11zdagtvjeDXOCPVXR0vcPPLFen7YU8DVvdK4Z1gHzBE6tNstRhKsJvlQEzFRWFjI0qVL6d69O8OGDSM+Pj7WQxIi5iRkCVEVV1GwyL06h3OD7RmKD8IlT0D3SdG1Z7DE4TcnUOjy4Y1Qf7VqVz6PfLmeMo+Pp8Z2ZUTXpmHHKCDRLsuDovbl5+eTm5tL//79SU1N5e677yYpKSnWwxKizpCQJUQkp7LB88avYPbjYEsKtmdo1qP6c8q7t7uVhcIyT9gkmdaafy3fzbvzttE8xc7bk3vQNjW8aanRoEi2m2XvQVGrfD4fixcvZtGiRRgMBjIyMkhMTJSAJUQFErKEqMjvK6+/Cu+sHnqcB+Y/D2v+A2m94bI/R9mewQT2FEp9UOIOXx4scfn409SNzN98hKGdG/PIZekRm4jaTEYS7bI8KGrXtm3bmDZtGsePH6dr164MHz6cxMTw/mxCCAlZQoSKdoPn4kPB9gwHfoBeN8Gg+6Nrz2C2oa1JFLn8uHzhRfRbDhfz+8/WcaDAxW+HdWBS7xZhIUoBCTYzdossD4ra5Xa7+fTTT4mLi+MXv/gF7dq1i/WQhKjTJGQJ8aNoN3jeswKy7wNvWXD2qtOo6s8p797uM9opKPNG3B5n2roDPD99Ewk2E+/8oicXtkgOO8ZoCG7uHKnwXYia4Pf7Wb9+Pd27d8dqtXLdddfRuHFjTCb5+BCiOvJTIoTWwdkrr6v641b9Axa+BMktYML/QaMO1V+/vHu7SxspitCeweML8OfZm/l89T56tkzm6SszaOiwhl3Gagp2b5flQVFbdu/eTXZ2NocPHyYuLo4OHTrQrFmzWA9LiHpDQpY4v0W7wbOnFGY/BnnToN0wGPlccB/C6piswe1xPP6I2+McKHTy8OfryD1QzPX9WnHr4LaYDKGzVApw2EzEWeTHVdSO0tJS5syZww8//EBiYiI///nPad++fayHJUS9I+/a4vzl8wQDVnUbPOfvgK/vhuPbYOB9wS1yoplNsjrQFgeFTi9uX/hrLN12jMe/Xo8/oHlxfHcGdwrfMNeggsuDFpMsD4raobXmP//5DwcPHmTAgAFcfPHFWCyWWA9LiHpJQpY4P0W7wfPWuTDjoWBR+7j3oVX/6s8p797uN1jIL/WE1V8FtOaDRTt4/7sdtGvs4Llx3WjZIC7sMpbyzZ0N0r1d1IIDBw7QqFEjzGYzI0eOxGazkZoaHvyFENGTkCXOL9Fu8Bzww5LXYcUUaJIR3B4nMYpaFKMF7Ml4AoqCUnfYTYqFZV6e+HoDS7cfY3S3pjw0snPEJqLxVlPEtg1CnG0ul4tvv/2WnJwcBg8ezODBg2nRokWshyXEOUHexcX5I9oNnp35MO1+2LUEMq6GoY8Ga6uqY4kHWyJOj59iV3iB+8b9RTz8+TqOlbr5/ajOXHlhs/D2DAqS7GasJmnPIGqW1pp169Yxa9YsysrKyMrKom/fvrEelhDnFAlZ4vwQ7QbPh9YHt8cpPQKX/gm6XV39tZUh2O3dbKPI5cXpCX0NrTVfrN7Hq7M308hh5b3rs0i/ILx5o7l8eVA2dxa1Yfbs2SxdupRmzZoxefJkuWtQiBogIUuc+6Ld4Hn9ZzD3SYhrCD//DzTtVv21y5cHtTJQUOrBU2H/QbfPzwvT88hed4B+bRvy5BVdSYozh10mzmIkwRb+uBBnk8fjwefzERcXx4UXXkiDBg3o2bMnBoPcWCFETZCQJc5t0Wzw7PPAvKdh3cfQsh+MfgXiGlR/bUscWBPxayIWuB8v9fDQZ2tZu7eQmwe24VeD2mCIsDyYaJPNnUXN0lqTl5fH9OnTadmyJePHj6dx48Y0btw41kMT4pwmIUucm6Ld4Ln4QLA9w6F1wdYMA34bbB5alZOWBz2+AAXO8A2etx4u4YFP1nC81MOzV2VwSXqTsMvI8qCoDfn5+UyfPp0tW7bQuHFjevfuHeshCXHekJAlzj3RbvC8e2lwexy/B8a8CR0urf7aRjPYU8BgLC9w94YVuC/acpTHvlpPvMXEX6/rFbH+ym4xkmCVzZ1FzcrLy+PTTz/FYDAwfPhw+vTpg9Eos6ZC1BYJWeLc4nWV119V0WBUa1j5Piz+MzRoG2zP0KBt9dcuXx5EqUoL3P+zYjdvzt1Kp6YJvHR1dxon2EKOUUCiXZYHRc3yer2YzWaaN29ORkYGQ4YMITExPOwLIWqWhCxx7nCXBDd5ru6YmQ/D1tnQcSQMfybYeqEqSpUvD9oJBDSFZeEF7l5/gBdn5PH1mv0M7dyYJ8Z0CQtSpvLNnU2yubOoIUVFRcyaNYuioiJuuukmHA4HY8eOjfWwhDhvScgS5wZXIXjKqj7m2Db4+i4o2A0XPwi9bqp+exyjGWzJYDTh8wcocHrDCtwLy7z8/vO1rNpdwE0DWnPLxW3DCtxtZiOJNlkeFDUjEAiwfPly5s+fj9/vZ9CgQQQCAVkaFCLGJGSJ+i+agLV5Bsz8A5hsMOEDaBFF00WzPTiDpRRun59CpzeswH3H0VIe+GQNh4vcPDW2KyO6Ng27TIJs7ixqUH5+Ph999BGHDh2iffv2jBo1igYNorg7VghR4+SdX9Rv1QWsgA++exW+/wAuyITLX4eE8CAU4qTlQaDSAvdl24/xhy/WYTUZeecXPenWPCnsMtK9XdQUrTVKKRwOB3a7nYkTJ9K5c2eZLRWiDpGQJeqv6gJW2TGYei/sXQGZ18Dgh8FkqfqaBlPw7kFj8EcjUoE7wCc5e3h19mbapjp45epMmiaFFrgbDYpkqb8SNUBrzapVq/j++++56aabMJvN3HDDDbEelhAiAglZon6qLmDt/wGm3gPOAhjxHHS9qvprnrQ8GAhoCp3esAJ3nz/Aq7M389mqfQzq0IinxnYNWwo0Gw0k280YpP+VOMsOHDhAdnY2+/bto1WrVjidTsxm2SlAiLpKQpaof6oKWFrD2o9g3jPgaAzX/Bcad6n6ekoFWzNY4gAqLXAvcnp55Iv1rNh5nOsuasXtP2sX1khUCtxFTfD5fMyePZuVK1cSFxfHVVddRbdu3eT7TIg6TkKWqF+qClheF8z9I2z8EloPglEvgT256usZTMFjjMHZgMoK3HcfL+P+j9ewv8DJY5enc3n38M10HVYT8Vb5kRJnn9Fo5NChQ2RlZTF06FBsNlv1JwkhYk4+EUT9UVXAKtwL39wNhzdC3zug353Vb49jtgXbM5TPBpR5fJS4fGEF7jk7j/Pw5+tQSvHW5B70aJkS8rw0GBU14ciRI8ydO5fLL78ch8PBddddJy0ZhKhnajRkKaVGAq8DRuB9rfXzlRw3AfgE6K21zqnJMYl6qqqAdSQPPr0JAl4Y+w60G1r1tZQCa0JIE9LKCty/WL2Pl2bm0bJBHK9cnUnzFHvI8walSI4zY5YCd3GWeL1eFi5cyJIlS7BYLBw+fBiHwyEBS4h6qMZCllLKCLwNXArsBVYqpb7WWm+scFwCcDewvKbGIuo5V1HlAetwbjBgmazw839Vvz1OheXBygrc/QHN63O38NHKPfRr25Cnr8zAYQv9cTEZFMlxFtngWZw1eXl5TJ8+ncLCQjIzM7n00kuJj69mRwIhRJ1VkzNZfYCtWuvtAEqp/wFjgY0VjvsT8CLwQA2ORdRXriLwlEZ+7tAG+PSXwbsCr/4HpLSq+loVlgcrK3Avcft49Mv1LN12jEm9W/CbS9pjMoTOVNlMRhLtUuAuzq4NGzZgtVq58cYbadWqmu9nIUSdV5Mhqzmw56Sv9wIhbbaVUj2AFlrrqUopCVkiVFUB6+A6+OxXYHEEA1Zyi8qvE2F5sLIC9335Tu7/ZA27j5fx0MhOjOuZFna5eKsJhxS4i7PA5/OxZMkSOnXqRJMmTRg9ejRms1mWBoU4R9TkJ0WkX/FPfKQppQzAn4Ebq72QUrcAtwC0bNnyLA1P1GlVBawDa+Dzm4NtF67+BySFB6ETDMby5qI/9RIq8/godvnCDl29O5/ff7aOgNa8MelCslqHbk0iBe7ibNq+fTvTpk3j2LFjaK1p0qSJ3DUoxDmmJkPWXuDk6YU0YP9JXycAGcD88iWXpsDXSqkrKha/a62nAFMAsrKyKt78Jc41VQWs/avg81+DvUEwYCWGt1I4wWQNLg+etNRXWYH71LX7eW7aJpol23nl6kxaNowLeV4pSLZbsJikwF2cmeLiYmbOnMmGDRto0KAB1157Le3bt4/1sIQQNaAmQ9ZKoINSqg2wD5gETP7xSa11IdDox6+VUvOBB+TuwvNcVQFr3/fBgBWfGgxYle1BqFRwGdHqOPFQVQXu787fxj+X7aJ36xSevaobifbQDtpGgyJFCtzFWfL999+zadMmfvaznzFgwABMJll6FuJcVWM/3Vprn1LqLmAmwRYOH2itNyilngJytNZf19Rri3qqqoC1dyV8cWuwi/uEf0BCk8jHGYzB2auT9iisrMC9zOPj8a828N2Wo4zr0Zz7h3cM22vQYjSm91J6AAAgAElEQVSQHGeWAndxRvbs2UMgEKBVq1YMGDCA7t2706BBg+pPFELUazX6K5TWehowrcJjj1dy7M9qciyijqsqYO1eBl/eDokXwIS/B4NWJBGWBysrcD9Y6OKBT9aw7UgJDwzvyIReaWFBym4xkmiTfeHE6SsrK2POnDmsXr2aVq1aceONN2I2myVgCXGekHlqEXtVBaxdS+CrO4LF7RP+DvGNIh9niQdbYshDlRW4r9tXyIOfrsXt8/Pnn1/IRW0bhh2TYDOFbfwsRLS01qxatYq5c+fidrvp378/gwcPjvWwhBC1TD5FRGxVFbB2fgdf3RXsfzXh/yAuPAwBwdora0LIQ5UVuM/ccJCnp+aSmmDl7ck9aJvqCHleKUiym7Ga5A5Ccfo2bdrE1KlTadWqFaNHj6Zx40pmX4UQ5zQJWSJ2qgpYOxbC13cFO7hP+L9gG4ZIrAlRFbgHtOa9hdv5YPFOerRI5vnx3UiOs4QcY1CKlDhzWF2WENFwuVwcOnSIVq1a0blzZyZNmkTHjh2lnk+I85iELBEb7uLKA9b2ecHNnht2gPF/qzxg2RJDGoxWVuDu8vp58puNfLvpMGMyL+ChkZ3D9ho0Gw0k280Y5A5CcYq01qxfv55Zs2bh9/v57W9/i8VioVOnTrEemhAixqIKWUopC9BSa721hscjzgfuYnCXRH5u61yY+ltI7RQMWLakyMdVCFiVFbgfLnbxu0/WknewmLsvac/kPi3DZhZsZiOJNtkiR5y6o0ePMm3aNHbs2EGzZs247LLLsFgs1Z8ohDgvVBuylFKXAa8CFqCNUupC4Amt9VU1PThxDqoqYG2ZBdn3QeMuMO79sEL2E2xJYPmpWWhlBe65B4r43SdrKfX4ePnqTAZ2CC+ad1hNxMsWOeI0FBQU8Je//AWz2czo0aPp1asXBoMsNQshfhLNp8tTBPccnAegtf5BKSXticWpqypgbZ4B2fdD024w7r2wQvYT7MnBDaHLVVbgPjf3EE9+s5GUOAtTru9Fh8ah15MtcsTpOnLkCKmpqSQnJzNy5EjS09OJj4+v/kQhxHknmpDl1VoXVFhKka1txKmpKmBtyobpD8IFmTBuSrBbe0VKBXtgmYN7u1VW4K615v8W7+SvC7fTrXkSL4zvRkOHNeQYg1Ikx5nD6rKEqEp+fj4zZsxgy5Yt3H777aSmppKVlRXrYQkh6rBoQlauUmoiYCjfIuceYFnNDkucU6oKWLnfwIyHoFlPuOqvIXVWJ1QIWJUVuLt9fp7JzmXmhkOM7NqUP1zWOawVg8mgSD6DLXICOkCRu4hEayIGJSHtfODz+Vi6dCkLFy5EKcWwYcOkmagQIirRhKy7gMeBAPA5wW1yHq7JQYlzSFUBa8MXMPMP0KIPXPkumOPCj1EqeHehKTgbVVmB+7ESN7/7dC0b9hdx++B23NC/VVghu9VkIMl++lvkeANeCt2FBHSg+oPFOcHv9/Pee+9x+PBhunTpwogRI0hMrKRWUAghKogmZI3QWj8EPPTjA0qpcQQDlxCVqypgrf8MZj0KLfvB2LdD6qxOUArsDU7sQ1hZgfvmQ8U88MkaCp1enh/XjSGdwxs/xlmMJJzBFjkun4tiTzFaVsrPC06nE7vdjtFopGfPnjRs2JD27aUUVQhxaqJZ73g0wmOPnO2BiHNMVQFr7ccw6xFo1R/GvlNJwDKEBKwilzdiwFq4+Qi3fPg9AQ1/va5XWMBSQKLNfEYBq8RTQpGnSALWeSAQCLBs2TJee+01tm4Ndqzp27evBCwhxGmpdCZLKTUCGAk0V0q9etJTiQSXDoWIrKqAteZ/MPeP0PpiuOLNE8uAIZQB4hqA0Vxlgfu/lu3m7Xlb6XxBAi9NyCQ1IfRaSkGy3YLFdHq1Uz/WX3kCnpDHtxVsI/dYLoNbyF5055I9e/aQnZ3NoUOHaNeundRdCSHOWFXLhYeB9YAL2HDS48XA72tyUKIeqypgrf4XzHsa2gyGMW+emKUKcVLAqqzA3eML8PyMTWSvPcCw9MY8dnmXsFYMRoMi5QwK3Curv1qwdwEvrniRJvFNGNB8ACaD9Ng6F8yePZslS5aQkJDA1VdfTXp6ujSnFUKcsUo/IbTWq4HVSql/a61dtTgmUV+5SyoPWKs+hPnPQrtL4LI/Rw5YBmNwidBowusPkF/mCStwzy/18PvP1/HDngJuHtiGmwe1CfswtBiDBe6nu0VOpPqrgA7w9w1/59+5/6ZLwy68MeQNCVj1nC7/5lJK0ahRI/r168fgwYOxWiPMrgohxGmI5lOiuVLqGaALYPvxQa11xxoblah/3CXBWaxIvv8/WPACtL8ULnsFjFUHLI8vQIEzPGBtP1LC/Z+s4Wixhz+N7crwrk3DLmO3GEk8w/qrMl9Z6GPeEp5b/hzLDixjVJtR3N3jblLjUk/7NUTsHTx4kOzsbDIzM8nKyqJHjx6xHpIQ4hwUTcj6O/A08DIwCrgJqckSJ6sqYK18H757GTqOhFEvgTFCADIYIa4hGIzBFg1l3rAS8yXbjvLol+uxmYy8+4ueZDQP39MwwWYiznJ6s0uV1V/tKd7DY4sfY3/Jfu7ucTdXtLtClpHqMbfbzbx581ixYgV2ux2bzVb9SULUQd9//31jk8n0PpBBdDexibMvAKz3+Xw39+rV63CkA6L5RIrTWs9USr2std4GPKqU+u6sDlPUX1UFrOV/gcWvQafRMOpFiLS8ZjAFa7AMRlxeP0XO0IClteajlXt4fe4W2jd28PLVmTRJDP1gVEBSnDms8Wi0Kqu/Wn5gOc8sewaz0cxLg18iMzXztK4v6obNmzfzzTffUFJSQlZWFkOHDsVuj3BnqxD1gMlker9p06bpqamp+QaDQW59joFAIKCOHDnS5eDBg+8DV0Q6JpqQ5VbBX923KaVuA/YB4Y2IxPmnqoC19G1Y+iZ0HgMjn6siYDUEgwGnx0+RyxvytM8f4OVZm/li9T4Gd0zlj1d0CZupMihFSpwZ02lukROp/kprzX83/ZcP1n9Au+R2PDXgKZrENTmt64u6w2g0kpCQwKRJk2jevHmshyPEmcqQgBVbBoNBp6amFh48eDCjsmOiCVn3Ag7gbuAZIAn45dkZoqi3KgtYWgfD1bJ3oMtYGP5scDmwIqM5WINlMERsMlri9vHwZ+tYsfM4N/RvxW2D22GosExnNhpIPoMC90j1V06fk5dXvsz8vfMZ2mIo92fdj83008yZQhFvjpctdeoBr9fLd999h1KKIUOG0K5dO9q2bSvLveJcYZCAFXvlfweVfiBUG7K01svL/7MYuA5AKZV2VkYn6qeqAtaS14PLhF3Hw6VPVRuwStw+St2hAetoiZt7P/qBbUdKeezydC7v3izsEjaTkUS76bQ+MCurvzpYepDHFz/O9sLt3NL9FiZ2nBhyfYMykGRJwhyprkzUKXl5ecyYMYOCggJ69OiB1hqllAQsIUStqjJkKaV6A82BRVrro0qprgS31xkKSNA6H1UVsBa9Civfg25Xw7Angz2vKjJagjVYSlHk8uL0+EOe3n2sjLv/t5qCMi+vXJ1Jv3YNwy4RbzXhsJ5egbs34KXIXYRfh77u6sOreWrpUwR0gGcHPUufpn1CnjcZTCRZkjBGCo2izigsLGT69Onk5eWRmprKDTfcQOvWrWM9LCHOSUajsVeHDh2cfr9ftW/f3vnxxx/vTEhIOOMb4/bs2WO6/vrrW+/fv9/i8/lUWlqae8GCBVubN2/ebdq0aZszMzPdPx77y1/+skWzZs08F110Udk111zTLi0tzeN0Og2NGjXy3n///QevueaawjMdz5modIpLKfUc8G/gWmCGUuoRYB6wBpD2DeejqgLWwheDAav7pMoDlsl6ImAVOsMD1ob9hfz6wxxcXj/vXNszLGApIMluPu2A5fK5KHAVhAQsrTWfb/mcBxc+SIothbcveTssYFmNVlKsKRKw6gGPx8OuXbsYNmwYt956qwQsIWqQ1WoNbNq0aeOWLVs2mM1m/corr0Td28br9Vb63EMPPdR86NChRXl5eRu3bdu24cUXX9wHcOWVVx7/8MMPT2zF4Pf7yc7OTrn++uvzAbKyskpyc3M37ty5c/0bb7yx+4EHHmj51VdfJZzBH/GMVVVYMhbI1FpfDQwHHgMGaa1f0VqXVXGeOBdVFbAWPB/shXXhtXDJE5UHLHsKGigo8+DyhgasJduOcse/VxFnNTLl+iy6NEsMed6gFCnxlrDO7tGKtP+gx+/hpZyXePuHt7nogot4a+hbpCWETtDGm+NJsibJMlMdtmPHDubMmQNAamoq9957LwMGDMBolFAsRG0ZOHBgydatW615eXmWDh06dP3x8ccff7zJfffd1wygT58+ne66667mvXv37vT000832b9/v2nEiBHtMjIy0jMyMtJnzZoVD3Dw4EFzixYtTtRz9O3b1wlw/fXXH//iiy9OhKzp06cnpKWluTt27Bha+wH079/f+bvf/W7/W2+9FdMb9aqaEnBprZ0AWuvjSqlNWuu8WhqXqEs8pZUHrHlPww//hh7Xw88eDm4YWFFIwArfhzB77QGeyc6lfWMHf/55Jg0doR23TQZF8mlukVNZ/dVR51GeWPIEm45v4vou13Ndl+tCitkVikRrIlajdP+uq4qLi5k1axbr168nJSWFAQMGYLfbsVgiNLsV4hz2u0/XtNh8sDjubF6zY9OEspcmZO6J5liv18vMmTMThw8fXlTdsQUFBcaVK1fmAYwZM6bNfffdd2jEiBElW7ZssYwYMaLD9u3bN9x5552Hb7zxxrbvvvtu2c9+9rOi22+//Vjr1q29ffv2dRoMBpYuXWrv16+f8z//+U/KhAkTjlf2Wn369Cl74403wrtW16KqQlZbpdTn5f+tgNYnfY3WelyNjkzUDV4nuCL83OgAfPsnWPNf6HUTXPxglQEroKHA6cV7UsDSWvPh0l28M38bvVun8Pz47mFLgVZTcIuc05lJ8gV8FLoLw+qvNhzbwB+X/JEybxlP9n+Sgc0HhjxvUAaSrEmYDVLgXhcFAgFWrlzJvHnz8Pl8DB48mAEDBmA2y9+XELXJ7XYbOnfu3AWgb9++xffcc8/RXbt2VfmDeM0115wIRYsXL07csmXLiWZ1JSUlxvz8fMP48eOLBg4cuO6LL75ImjFjRlKvXr26rFu3bkOzZs1848aNO/6vf/2rQVZW1r7Zs2cnv/TSS/srey1dcduQGKgqZI2v8PVbNTkQUQf5POCKUDOoAzDnCVj3CfT+NQy8L3LAMtuCASugyS/z4Dtpo2d/QPPanM18nLOX4V2a8PiYLpgr9LqKsxhJOM0tciL1vwKYtmMab6x6g0b2Rrx48Yu0SWoT8rzFYCHRmigtGuowl8vF/PnzSUtLY9SoUTRsGH5zhBDnk2hnnM62H2uyTn7MZDLpQOCnX6ZdLlfIm+nJhfFaa3JycnIdDkdYGmrSpIn/tttuO37bbbcdHzJkSPtZs2Y5brzxxoIbbrjh+MiRIzsMGTKkuFOnTs7mzZv7Kp77o5UrV8a1b98+pnsvV/pJorWeW9U/tTlIEQN+HzjzCdtAUAdg1mPBgNXn1ioClh3sKfgDmuMVApbHF+CxL9fzcc5eJvdpyZNju4YFrASb6bQDVqT6K1/Axxur3uCVnFfITM3knWHvhAUsu8lOsi1ZAlYdVFZWxsKFC9FaExcXx6233sq1114rAUuIOiYtLc13/Phx08GDB41Op1PNnDkzfA+0cgMHDix64YUXTtRMLVmyxA7w9ddfJxQXFxsA8vPzDbt27bK2adPGA9C1a1d3cnKy/9FHH02bOHFipUuFy5cvt7/00kvN7rzzzojb3dSW07tNS5zbAn5wHg8GqoqPz3oENn4JF90B/X4TOWBZ4sCWhM8fIL/MS+CkoFbi8vG7T9ewancBd1/Snmv7tgo59Uy2yKms/qrAXcBTS59izZE1TOw4kZu73Rxyp6BC4bA4sJtki5W6RmvN6tWrmTNnDi6Xi7Zt25KWlkZycnKshyaEiMBqter777//QJ8+fdLT0tLcVc0kTZkyZc/NN9/csmPHjl38fr/q27dvcf/+/XevXLky7t57721pNBq11lpdd911RwcPHnzihrsJEyYce/bZZ9OuvfbagpOvl5OT40hPT+/idDoNDRs29L700ku7x44dW8m2JLVD1YU1y1ORlZWlc3JyYj2Mc5fWUHYM/BVurw34YMbDsOmbYLjqd2fk8y3xYEvE6w+QX+YJmQg7Uuzmtx/9wI6jpTx+eRdGZoTWIxqUIjnOHDarFY3K6q8252/miSVPUOAq4P6s+xnWaliF15QGo3XVwYMHyc7OZu/evbRs2ZLRo0fTpIlsbyTOP0qp77XWWSc/tmbNmp2ZmZlHYzUm8ZM1a9Y0yszMbB3puahnspRSVq21u/ojRb2ldXCJMFLAmv4g5E2DAb+FvrdFPr88YHl8AQrKPCHVULuOlXLP/36g0Onl1YmZXNQ2dJnnTO4grKz+6tvd3/JyzsskWhJ5fejrdEwJbe8mDUbrLq01n332GWVlZYwdO5bMzExpoyGEqHeqDVlKqT7A3wjuWdhSKZUJ3Ky1/k1ND07UMlcB+Crk6IAPpj0Am2fAoAeg982Rz7U6wJqAy+unyOkNiTvr9hVy/8drMCh459qepF8Q2gPrTO4gjLT/oF/7+du6v/FR3kdkNMrgiX5P0MDWIOQYq9FKoiVRPrjrEK01ubm5tG/fHovFwoQJE0hMTMRul2VcIUT9FM1M1hvA5cCXAFrrNUqpITU6KlH7XEXgrbB0rjXMfiwYsC5+ELIq2RfcmgBWBy6vn0Jn6CzYoq1H+cPn60hNsPL6pAtJSwlt5WK3GEk8jQL3yuqvij3FPLP8GVYeXMkV7a7gjgvvCGvF4DA7iDOf1ZYy4gwdO3aMadOmsX37dkaMGMFFF10kS4NCiHovmpBl0FrvqvAbv7+yg0U95CkN/nOyH7fK2fAF9Lur8oBlSwRLPGUeH8Wu0Dtpv16zn+enbaJDEwevTgxvMuqwmog/jS1yKqu/2lm0k8cXP86h0kPc1+s+Lmt7Wcjz0mC07vF6vXz33XcsWbIEk8nEqFGjyMrKqv5EIYSoB6L5hNtTvmSolVJG4DfA5podlqg1XlfkZqMr3/9pq5yLKilytyWBJY5St48S908BS2vN35fs5C8LttO3TQOeG9ctJEwpINFuPq0tctx+N0XuorD6q8X7FvPciuewGW288rNXyGiUEfK8URlJsiZhMsgNtXXJN998w7p16+jevTuXXnopDocj1kMSQoizJppPnNsJLhm2BA4Bc8ofE/WdzxOsw6po3Sew6BXodBkMeSRym4bygFXs8lJ20kbP/oDmlVl5fLZqHyO7NuXRy9ND7hZUCpLtFiymU7+DMFL9VUAH+Ffuv/jHhn/QKaUTT/Z/ktS40D1KpcFo3VJQUIDJZMLhcDBo0CB69uwpGzkLIc5J0YQsn9Z6Uo2PRNSuypqNbpkV7ObeehCMfC7yZs/2ZDDbKXR6QzZ6dvv8PPHVBublHeHavi25a2h7DCcFNKNBkXIadxBqrSnyFOH2hxbll3nLeGHlCyzat4hLW13Kvb3uDVsKjDPF4bDI7Ehd4Pf7WbJkCQsXLqRr165ceeWVpKamkpqaWv3JQog6affu3aY77rij5Zo1a+IsFotOS0tzv/nmm3u6d+8e1o0gLy/PkpmZmdG6desTBcA//PBD7pQpUxo88cQTaU2aNPECpKenl33xxRc7a/GPUWOiCVkrlVJ5wEfA51rrmDb2EmdBIBC52ejuZTDtfmjaHca8DsYKG+0qBbZktMlKUZkXl++ngFXs8vK7T9ayek8B91zSgcl9W4acajEG7yA0nGLAqqzAfX/Jfh5b/Bi7i3dzR+YdjOswLuROQWkwWrfs2LGDadOmcfToUdLT0xkyRO6dEaK+CwQCXHHFFe0nT558bOrUqdsh2LV9//795kghC6BFixbuilvxAIwZMyb/ww8/3F3TY65t1a6faK3bAU8DvYB1SqkvlVIys1VfaR0MWIEK9y4cWg9f3QHJreHKv0DFu++UAnsK2mSloELAOlzs4tZ/fs+6fYU8NbZrWMCymY0kx516wPIH/BS4C8ICVs6hHO6YcwfHXMd4YdALjO84PiRgGZSBZGuyBKw64vvvv+fDDz/E7/czefJkJk6cSFJSpTttCCHqialTpyaYTCb94IMPHvnxsf79+zuHDx9ecuutt6Z16NCha8eOHbu89957Kadz/Q0bNlgHDRrUoWvXrum9evXqtHr1ahvA/v37TSNGjGiXkZGRnpGRkT5r1qz4s/VnOtuiqgLWWi8Bliil/gi8Bvwb+F915ymlRgKvA0bgfa318xWevw24k+DdiiXALVrrsIQrzqJIzUaPb4fPfx1cBhz3fvDfJysPWAGDhYIyL17/TzNgO46Wcs//VlPs8vHnn19Inzah/ajirSYcp3kHYYG7gMBJs21aaz7Z/AnvrX2PVkmteKr/UzRzNAs5TxqM1g2BQICysjIcDgedO3empKSE/v37YzZLZ30hzrov72zB4Y1nty9N4y5lXPl2lRtPr1271p6ZmVlW8fEPP/wwed26dfbc3NwNBw4cMPXp0yd9+PDhJQB79uyxdu7cuQtA7969S/75z3/uBvjmm29SOnfu7AC4/fbbD91zzz3Hbr755lZTpkzZ1a1bN/e3334bf/vtt7dctmzZ5ltvvbXFfffdd2jEiBElW7ZssYwYMaLD9u3bN5zVP/9ZEk0zUgcwFpgEpANfAf2jOM8IvA1cCuwluOz4dYUQ9R+t9V/Kj78CeBUYeap/CBElZ4Rmo8UH4bObAQXjP4CECr2JTgpY+RU2el67t4D7P16DyWjgL7/oRaemCT+dxunfQej1eyn0FIYELLffzSs5rzB391wGNR/EQ30eCpupkgajdcPevXvJzs7GZDLxy1/+kvj4eAYPHhzrYQkhasl3332XMHHixOMmk4kWLVr4+vbtW7Jo0aK4rKwsZ7TLhYWFhYbVq1c7rr766nY/PubxeBTA4sWLE7ds2XLiA6CkpMSYn59vSElJqVADE3vRTDGsB74BXtRaf3cK1+4DbNVabwdQSv2PYFg78T9Xa31y74B4oH5tpFifuIvB6wx9zJkPn98M7kKY+E9IaR36fHnA8pcHLP9JAWvh5iM8+uV6UhOsvDGpB81T7CGnne4dhJFaNBxzHuOxxY+xOX8zv8z4JZM7Tw4LUtJgNPacTidz5sxh1apVJCQkMGLEiFgPSYjzQzUzTjWlW7duzi+//DJsKfBs7Ins9/tJSEjwRQpkWmtycnJyHQ5Hnc8M0XwKttVa/+YUAxZAc+Dkv/i95Y+FUErdqZTaBrwI3B3pQkqpW5RSOUqpnCNHjkQ6RFTFUwbukgqPlcIXt0HBbhj7DjTuEn6eLRmfMnO8NDRgffXDPh76bC3tUh28d31WSMAyGhQN4k4vYDl9TgrdhSEBa1vBNu6aexe7inbx1ICnuDb92oj1VxKwYuvAgQO89dZbrF69mosuuog777yTrl27yqyiEOewMWPGFHs8HvXKK680+vGxBQsWxKWkpPg+/fTTBj6fj/3795tWrFjhGDRoUGlV16qoQYMGgbS0NM8HH3yQAsEShKVLl9oBBg4cWPTCCy80/vHYJUuW1NkC3EpnspRSr2it7wc+U0qFpUWt9bhqrh3p3TXSdd4G3lZKTQYeBW6IcMwUYApAVlZWnU+udYrXBa7C0Mf8Hvjmbji0Dsa8AS36hp9nT8ZnsHC8zHOiy4PWmg8W72TKwu30a9uQZ8dlEGf56VvIbDSQfBp3EAKUeksp9Yb+DC4/sJw/LfsTcaY4XhvyGh1SOoQ8Lw1GY8/n82EymUhNTaVdu3YMGDBAtsMR4jxhMBj4+uuvt91xxx0tXnvttaZWq/VEC4eSkhJjenp6V6WUfvLJJ/e2bNnSl5eXZ6n+qj/573//u/3Xv/51qxdeeOECn8+nrrrqquP9+vVzTpkyZc/NN9/csmPHjl38fr/q27dvcf/+/evknYmqsmk9pVQfrfUKpdQlkZ7XWs+t8sJK9QP+qLUeUf71w+XnPVfJ8QYgX2td5W1HWVlZOicnp6pDxI98nvJWDSf9HQf8MP13kDcNhj8DGePDz7Ml4TPayC/zEig/1x/QvDwzj89X72N0t6Y8Mjod00lNRm0mI4l202nNXBR5inD5QvdN/HLrl7y9+m3aJrfl6YFPk2qXBqN1idvtZv78+WzevJnbbrtNCtqFqEFKqe+11iH7Ta1Zs2ZnZmbm0ViNSfxkzZo1jTIzM1tHeq7SKQCt9Yry/0zXWr918nNKqbuAKkMWsBLooJRqA+wjWDg/ucJ1Omitt5R/eRmwBXF2RGo2qjXMeyYYsAb9rpKAlYjfZCe/1HMiYLm8fh7/agMLNh/h+n6tuONn7ULCVJzFSMJpbPIcqcmoX/t594d3+WLrF/Rv1p8/9P1DWIG7NBiNHa01GzduZObMmRQXF9OrVy8CgTpXayqEEHVCNOssvwTeqvDYryI8FkJr7SsPYzMJtnD4QGu9QSn1FJCjtf4auEspNQzwAvlEWCoUpyEQKA9YFT78lr4Ja/4DWb+C3r8KP8/qwG+KI7/sp4BV5PTywCdrWLu3kHuHdWBSn9AeWIk2M3bLqd9BGNABCt2FeAM/tZMo85bx9PKnWX5gORM6TuCW7rdgVD9dW6FIsCRgM9lO+fXEmXO5XHzyySds376dpk2bMnHiRNLS0mI9LCGEqLOqqsn6OcHZpzZKqc9PeioBiLDhXTit9TRgWoXHHj/pv+85pdGK6p1oNuoLfXz1P2HZO9B1PAx6IPw8SzwBsyPkLsJDRS5++78f2JNfxp+uzODSLj/V2iggKc6M1XTqAevHJqN+fVJD07LDPLLoEXYW7eSenmGWUIgAACAASURBVPdwRbsrQs4xKANJ1iTMBlmWqm1aa5RSWK1WTCYTo0aNIisrC4NBlmqFEKIqVc1krQCOAWkE+139qBhYXZODEmcgUrPR3G+Cy4TthsGlT4Zv+GyJI2BJCAlYBwqd3PHvVRSUeXnt5xeS1fqnJqMGpUiJM4fUZEXLG/BS6A7tgZV3PI9HFz+K2+fm2YHP0rtp75BzzAYzSdYkqb+Kgc2bNzNv3jwmT55MQkICkyZNkjsGhRAiSlXVZO0AdgBzam844oxEaja6YyHMfBjSesNlr0DFO/HMdrQ1kfzSnxqNHih0cvu/VlHs8vH25J50aZZ44nCTQZF8Gps8A3j8nrAWDYv2LeLZ5c+SbE3mxaEv0iapTcg5NpONBHOCfLDXsoKCAmbOnMmmTZto1KgRpaWlJCTI34MQQpyKqpYLF2itByul8gltvaAArbVuUMmpIhYiNRvdvyrYqqFRRxj7Lpisoc+bbWhbEvll3hMBa39BcAar2OXjzWt6hAQsqym4yfPpfNC6fC6KPcUnApbWmo83f8x7a9+jc4POPDXgKRrYfvqWUijizfHS/6qWaa1ZvHgxCxcuBOCSSy6hX79+GI2yTZEQQpyqqpYLh5T/u1EVx4i6IFKz0SN5wWajjiYw7j2wVrgbz2RF25JD9iI8OWC9NbkH6Rf8FLDsFiOJp3EHIQQL2ku8P43PF/Dxxqo3yN6RzeC0wTzU5yGsxp8CoEEZSLIkYTZK/VVtU0px5MgR2rVrx4gRI0hOTq7+JCGEOAvy8vIs8+bNc9x2223Ha+s133jjjYY5OTnxJ2/pczZVWuSi9YmimRaAUWvtB/oBtxLcAkfUBT53eLPRwr3B7XJMNhj/N4hrGPp8ecAqdHrxVAhYJe7wgJVgM512wPp/9u47Pqoq/eP458wkM8mkF1qoQRKSAAm9F6lSFFwBQRD7+lN3dXcVxYqIooggirpr2WWxseJaWQmoSBWkhN6lQySUhPRkMu38/pgAAQNMQnqe9+vFi8zMnXufOSLz5d5zn5Njy7koYOXYcnhq9VMsOryIcTHjeLbrsxcFLC+DFyHmEAlYFSgnJ4evv/6akydPAjB8+HDGjBkjAUsIUaH2799vXrBgQZW+Sma326++URGezCT+BtBKqeuAj3AvEj2/5KWJMue0uye6F5WbCl/e625EOvKfEHTJLfZGE/iGkGl1UOD4fcB667YLAUsBQb7eF3V195TWmsyCTPIcFxZoP5FzgoeXPcz2M9t5vNPj3Nvm3osms5uNZkLMIRgNcmmqIrhcLjZs2MDbb7/Nrl27SElJAZBLg0IIj+zbt88UGRnZasyYMU2joqJaDR8+PPKbb74JaN++fUzTpk1bL1++3HLq1CnjgAEDrouOjo5LSEiIWb9+vS/AokWL/GNiYuJiYmLiYmNj49LT0w3PPPNMw6SkJP+YmJi4F154oe6cOXPC+vfvf12vXr2imjVr1vqxxx5rcO7YU6ZMqRcVFdUqKiqq1dSpU+sWreeWW25pFh0dHTd48ODm2dnZBoCGDRu2SUlJ8QJYtWqVpXPnzi0v/Tzz588Pio+Pj4mNjY3r3r179PHjx70AHn300YjbbrutaY8ePaJuueWWyEvfdyWefHu6tNZ2pdQtwBta6zlKKbm7sLKd74VVZLpcQTZ8fT/knIZRc91zsYoymsASSma+hwGrlC0azgUsm8t2/rldqbt4bs1zuLSLV3u/Stu6bS96jyzwXLF+++03Fi1aREpKCs2bN2fo0KGEhYVd/Y1CiCrnuTXPNT6QfqBM/wJtEdIi78UeL1514enjx4/7LFiw4FCHDh2OxsfHx3766adhSUlJe+fPnx88bdq0Bg0bNrQlJCTkLV269ODChQsD7rzzzsi9e/funjVrVv05c+YcHTRoUG5mZqbBYrG4pk2b9tusWbPqLV++/AC4L+Vt377db8eOHbv8/f1d7dq1ixsxYkSmUor58+eHbdq0aY/Wmg4dOsT2798/Ozw83HnkyBGf995778igQYNyR48e3ey1116rM3Xq1FOefOaBAwfmjB07dq/BYOD1118Pnzp1av0PPvggGWD79u2W9evX7y3potSenMlyKKVGAxOA7wqfk2s5la0g071EzjmOAvj2IUj91b0eYUS7i7c3ep8/g2V1uN93IsN9F2HupQFLQbDFVKqA5dIu0gvSLwpYy44t47GVj+Fv8uet/m9dFLBkgefK8euvv5KTk8OoUaO4/fbbJWAJIUqlYcOGBZ07d843Go1ER0fn9+vXL8tgMNC+ffu85ORk84YNGwLuvffeNIDhw4dnZ2RkeKWlpRm7du2aM3HixMYvvfRS3dTUVOPllubq2bNnVv369Z3+/v562LBh6StWrPBfsWKF/9ChQzMCAwNdQUFBrmHDhqUvX748AKB+/fq2QYMG5QJMmDAhbe3atR4vD3L48GFTr169oqKjo+PmzJlTf+/eveeXGxk8eHBGSQMWeN7x/SFghtb6UOEyOf8p6YFEGbLluRd+PsflgEWPQnISDH0NIntdvL3BC3xDybI5sdovDlh5NgdvjWtHTP0LASvEYsK7FD2wHC4HGQUZ53tgaa35ZM8nzNs1jzbhbXih+wsEmS8sTSkLPFccrTVbt24lICCAFi1a0LNnT7p3747ZbL76m4UQVZonZ5zKi8lkOh88DAYDPj4+GtzTDpxOpzIajb8LJkop/fLLL5+8+eabM7/99tug7t27xy5ZsuTX4vZ/6d3sSikut+by5bYvrEefWwIsPz+/2C+4P//5z03+8pe/nBw/fnzmd999FzB16tSIc6/5+fmVav2wq36Taq13Ao8ASUqpGOC41npaaQ4myoDTAQVZFx5rDT9OhoM/Qd9nIObGi7c3GMESSrbNSb7NHbB+Sy8+YBmUIrSUAcvutF8UsGxOG9M3TGfernkMbDqQGb1nXBSwzEYzoT6hErAqwKlTp/j3v//NwoUL2b59OwDe3t4SsIQQ5a5r167Z//73v8MAvvvuu4CQkBBHaGioa9euXebOnTvnT5s27WSbNm1yd+7c6RMUFOTMycm56BLKzz//HHjq1CljTk6OSkxMDO7Tp09Ov379chITE4Ozs7MNWVlZhsTExJC+fftmA6SkpJiWLl3qBzB//vzQ7t275wA0atTItmbNGgvA559/HlJcrdnZ2cYmTZrYAebNm1cmp/ev+g2nlOoFfIx7kWcF1FdKTdBarymLAkQJWTMvnoe1eibs+gq6/gna3X7xtgYjWMLIsWvyigasTzeRb3P+LmCVtot7gbOArIKs8z2wMgsymbx2MjtTd3J3q7sZHzv+on9d+Hn74ectN6iWt4KCAlasWMH69evx9fVl+PDhtG3b9upvFEKIMvLqq6+eGDduXLPo6Og4X19f17x58w4DzJgxo+7atWsDDQaDjo6Ozh81alSmwWDAy8tLt2zZMm7cuHGpISEhzo4dO+aMGTMm8siRIz4jR45M6927dx7AuHHj0tq3bx8LMGHChDM9evTI37dvn6l58+bWuXPnhj300ENNIyMjCyZOnHgGYPLkySceeOCBZq+++qq9Q4cOucXV+swzz5y47bbbrqtXr56tY8eOuceOHbvmf4mqK512A1BKJQF3aK13Fz6OBT7WWne81oOXRseOHXVSUlJlHLryFWRf3A9r4z/dISthHPR77uLlcpQBLGHkOiCnwL2O4fmAZXfy9m3taVk/AACjQRFSyi7uVoeVLNuFM2vHs4/z9OqnOZN/hic6PUG/Jv0ulIQi0Bx4UcsGUX62bdvGN998Q/v27RkwYAC+vr5Xf5MQospRSm269Dt327ZtRxISElIrq6aKUNIeVvv27TPdeOONUfv3799V3rUVtW3btvCEhIRmxb3mybUa07mABaC13qOUMpVVccJDDtvFAWvfYnfAajkU+j1bTMAKJc95IWAlp+fx0KebfxewvAoDlqEUAavAWXBRwNpyegtT1k7By+DFrD6zaBXe6vxrMv+qYqSlpZGamkrLli2Jj4+nXr161K9fv7LLEkKIWsmTb7zNSqn3cF8yBBiPLBBdsVwusGZceHz2EPzwDDRoC4Onu0PVOUqBJZR8p4Fsq7tpWnJ6Hg9+shmrw8k749oTXa9sAlZmwYUmqEsOL+H1Ta/T0L8h03pOI8L//HxBTAYTgeZAWeC5HNntdn7++WfWrFmDv78/LVq0wGg0SsASQlRbjzzySBqQ5un2LVu2tFX0Wayr8SRkPYB74vsTuOdkrQLeKs+ixCWKtmuw58H//uJeh/DGN9y9r85RCnxDsboMZF0lYHkbDYRYSrcO4bk5WOBu2TB351z+s/c/tK/bnue7PY+/6cIdsxYvy0WPRdnbv38/ixcvJj09nTZt2jBw4EBpKCqEEFXAFUOWUqoNcB3wtdZ6RsWUJC5StF2D1rB0CqQdcHdzDyhylkIp8A3BhhdZ+e4eVcfPui8RXhqwTEYDwdcYsDQah8vBqxteZdnxZQyLHMYj7R85fzlQoQgwBeDj5XNNH19c2alTp5g/fz5hYWHccccdREaWqBmxEEKIcnTZkKWUehq4F9gMdFJKTdVaz62wysTv2zXs+Bz2LIRuD0PTHhdv6xOM02AiI7cAjTtgPfjpZmwO10UBy+xlIMj32gOWzWlj6i9T+SXlF+5tfS+3xdx2fp8GZSDIHIS3QXrWlgen08mxY8eIjIykXr16jBkzhqioKDl7JYQQVcyVzmSNB+K11rlKqTpAIiAhqyIVbddwaicsfwma9oSuD168nW8w2stMRq4NrS8OWG+Pa3c+YPl4GQn09brmgJXvyOf5tc+z6dQmHm73MDe3uPn8dt4Gb4LMQTL/qpwcOXKExMREUlNTefjhhwkJCSEmJqayyxJCCFGMK30TFmitcwG01meusq0oawXZ4Cxcmsaa6Z6HZQl3d3QvGmB8gsDbl8x8Ow6X5lSWlYfOncEaXyRgeRsJKoNLhDn2HJ5c/SRbTm3h8U6PXxSwfL18CTYHS8AqBzk5OXz99dd8+OGH2O12xo4dS0hIsf30hBCiWiq6iHNNcaUP01wp9VXhzwq4rshjtNa3lGtltVnRdg3aBUsmuRd9vvVj8C3yxWoOAJOFLKudAoeLbKudvy3YSk6Bg3dv70BUXXfA8jUZCfQp3aU7m9N2PmBlFmTy1OqnOJBxgKe7Pk3fxn0B9/wrf5M/vl7Sh6k82O123n33XfLz8+nVqxe9evXicut8CSFEZXC5XGitZdrCJa4UskZe8vjt8ixEFLq0XcPGf8GhFe4lcyKKdOv29gWzP3k2B/k2JzaHiye+2M6RtDzeGNP2fB8si8lIwDUErMyCTDSas9azPLHyCZJzkpnSfQrdI7oDhfOvTEF4G+VLv6ylpaURFhaGt7c3AwcOpGHDhoSHh1d2WUIIAbibfw4ZMiSqe/fu2Zs2bfJv1apV3t69e32tVqvhpptuSp89e/YJcJ+huvXWW9O+//77IIfDoRYsWHCoXbt21pMnTxpHjhzZ/OzZs97t2rXLLdocfcqUKfU+/fTTcHB3dJ88efLpffv2mQYPHhzVuXPnnM2bN/vHxsbm3XPPPalTp05tmJaW5jVv3rxDffv2zauk4SjWZUOW1vqniixEFCraruH4elgzG6KHQNsiS+YYTeATRIHDSbbVgUtrpn63m83HMpgyPI7OkaEA+Jm98DeX7sxr0YB1Ou80E1dOJC0/jWk9p9GhXgfAHbCCzcHSYLSM5efn89NPP7Fp0ybGjx9PixYtSEhIqOyyhBBV1Imnn2lcsH+/pSz3aY6Kyot4edpVF54+cuSIzwcffHDkk08+OXbq1CljvXr1nA6Hg+7du7dcv369b5cuXfIBwsPDHbt3794zffr0OtOnT6+3YMGCo08++WREt27dcmbOnJny2WefBf3nP/8JB1i9erVl/vz5YZs2bdqjtaZDhw6x/fv3zw4PD3ceP37cZ8GCBYc6dOhwND4+PvbTTz8NS0pK2jt//vzgadOmNejbt+/BshyHayWTZ6qSou0ack7DoscguCkMevFCR3eDEXxDcLg0mXnuXlhvLzvAj7tP8dD11zGkdQMA/MsoYJ3IOcFfl/+VDGsG03tPPx+wvAxessBzGdNas3XrVt5++202b95Mly5daNy4cWWXJYQQl9WgQQNb//79cwE+/PDD0Li4uNi4uLi4/fv3+2zbtu18D59x48alA3Tu3Dnv+PHjZoB169YF3HPPPWkAY8eOzQwMDHQCrFixwn/o0KEZgYGBrqCgINewYcPSly9fHgDQsGHDgs6dO+cbjUaio6Pz+/Xrl2UwGGjfvn1ecnJylVuzTb4hq4qi7RpcDlj0KNhyYdRcONfMs7DZqAtFep4NDSzYeJxP1x9jZPuG3NGtKeAOWH5lELCOZB3hiZVPYHPZmHn9TKJDogF3B/cgc1CpJtGLy/v888/Zu3cvjRo1YtiwYdKtXQjhEU/OOJUXi8XiAti7d6/p7bffrrdp06Y9derUcY4cObKZ1Wo9fyLHx8dHA3h5eWmHw3H+y8Ng+P25niutqWwymc6/aDAYzu/XaDTidDqr3JeSx2eylFJVLiHWKEXbNax5E35LggFTIDz6wjY+wWiDkYx8Oy6tWbb3NLN//JU+0XV4bFBLlFJYTMYyCVj70/fz6PJHcWkXs6+ffT5gmY1mCVhlyGaz4XK5AIiLi2P48OHcc889ErCEENVKenq60dfX1xUaGuo8fvy414oVK4Ku9p6uXbtmz507Nwzg888/D8zKyjIC9OvXLycxMTE4OzvbkJWVZUhMTAzp27dvdnl/hvJw1W9jpVRn4F9AENBEKZUA3Ke1fri8i6s1irZrOLgMNn4A8WMgbsSFbcwB4O1DVp4du9PFlmPpPP/tLlo3DGLqiFYYDQrfMprkvittF0+tfgo/Lz9e6/MajQIaAe4WDQGmgGv9tAL3v9T27NnDkiVL6NWrF506daJNmzaVXZYQQpRKt27d8lu3bp0XFRXVqkmTJgUdOnTIudp7pk+ffmLkyJHN4+LiYrt165bToEEDG0DPnj3zxo0bl9a+fftYcE9879GjR/6+fftMV95j1aOudFoOQCm1DhgDfKO1blf43E6tdesKqO93OnbsqJOSkirj0OXDYYO8wvUvM47DpyMhqBGM/Y97fUIAbx/wDSGnwEFugYNDZ3L4v483EWwx8c87OhJk8cbHy90HqzSKBqytp7fyzM/PEOoTysw+M6nnVw8Af29/LN5lOq+y1jp79iyJiYkcPHiQ+vXrM2zYMBo1alTZZQkhqiil1Catdceiz23btu1IQkJCamXVJC7Ytm1beEJCQrPiXvPkupJBa330kstDzrIorNbT+kK7BkcBfPcX9883zbkQsIwm8Akm3+Ykt8DBmewC/rZgG15GA2+ObUuQxRuT0UCgb+kuEdqd9vMBa33KeqasnUID/wa81vs1wnzDAAgwBUgPrDKSlJTEkiVLMBqNDB48mE6dOhU7J0EIIUT158k38/HCS4ZaKWUEHgZ+Ld+yaglrkXYNK16G07thxN/dZ7Lg/J2ENqcm22onp8DBXxdsJTPfzj9ub09EsC/e17DYs91pJ6MgA41mVfIqpq2bRrOgZszoPcM97wpFoDkQs1Gm410rl8uFwWAgNDSU2NhYBg0aRECAXHoVQoiazJOQ9SAwB2gCnAKWFj4nroU93/0LYM//YPsC6PRHuK6f+zml3K0aNGTk27A5XTz55XYOp+Yya3QCsQ0C8TIogku52HPRgPXj0R+ZsWEGMWExvNLzFfxN/tJktIxkZmby/fffExwczKBBg2jevDnNmzev7LKEENWfy+VyKYPBcOU5P6JcuVwuBbgu9/pVQ5bW+jQwtiyLqvVcTrAWtmtIPwpLn4eGHaDHXy5s4xOES3mRkWfD5dK8tGgPG4+k89yNsXS7LgyjQRFiMWEwXFvA+t/B//Hm5jdJqJPASz1fwtfLV5qMlgGn08m6detYuXIlWmuuv/76yi5JCFGz7Dxz5kxcnTp1MiVoVQ6Xy6XOnDkTBOy83Dae3F34AfC7/4Ba6/uvrbxazJrpXpPQYYNFfwODNwydCedCjdkf7eVDZp4dp0vz9xUHWbLzJP/Xuzk3xkdgUGUTsP776395d9u7dGnQhee7PY/ZaMbL4EWQKQijQdafKq2UlBS+/vprzpw5Q8uWLRk8eDDBwcGVXZYQogZxOBz3nTx58p8nT55sjTQWrywuYKfD4bjvcht4cqpiaZGffYA/AJXW+KzaK8hxT3IH+HmWex7W8LchwN2pHS8zmAPIyrdjc7r4YlMyH/1ylJvbRnB3j2YoBSEWb4zXELBc2sXHez7mw10f0rtRb57u8jTeBm9MBhOB5kAMSv5/vRbe3t64XC7Gjh1Ly5YtK7scIUQN1KFDh9PA8MquQ1yZJ5cLFxR9rJT6GPix3CqqyZx2sBW2Djm0HDZ/CG3HQ4sB7ueM3uAbQm6BA6vdycp9Z5j5/T56RYXz+OCWGJQi2NeEl7HkIahowHp/x/t8vu9zBjUdxMSOEzEajJiNZgJNgdJktBRcLhebNm0iJSWF4cOHEx4ezp/+9CcZSyGEqOVKM+kmEmha1oXUeFpDfob79+xTsOQpqBMDvZ9wv64M4BuC1eEip8DB9uQMnvt2J3ERgbw4ojXeBoO7XYNX6QJWpi0Tp3by1pa3WHhwIcOvG87D7R7GoAzSZPQa/PbbbyQmJnLixAkiIyOx2+14e5fuZgQhhBA1iydzstK5MCfLAJwFnizPomqkgiz3moQuJyye6O7wPuz1C/2wfEOwa0VWvo2jabk89t9t1AkwM2t0Ar4mI4G+3pi9Sj5Pyu5yByy7087MpJn8cPQHbo2+lfvj70cphZ+3H37efmX8YWs+q9XKTz/9RFJSEv7+/owcOZJWrVpJuBJCCHHeFUOWcn9jJAC/FT7l0ldrES9+z24FW5775/XvQvJGuOEVCC28ld8cgNPgTUaujdScAv7y2VaMSvHm2LaE+JkI9PHGx7uUAasgkwJnAa+sf4WVySu5s9WdTIidgFJKmoxeA5fLxe7du+nSpQvXX389Pj4+V3+TEEKIWuWKIUtrrZVSX2utO1RUQTWOy+W+mxDc4WrdOxA7HOJudj/nZUab/MjItZFttfO3z7eRnmfjH+M70CjEgr/ZC19T6QOW1WFlyi9TWJ+ynv+L/z9ubXmrNBktpdOnT7Nx40aGDBmCxWLhkUcewWyWMRRCCFE8Tyb4bFBKtS/NzpVSg5VS+5RSB5RSv7vEqJR6VCm1Wym1XSn1k1Kq5s31sma42zXkp0Pi4xDUGPpPdjcbNRjBJ5isfPdE96e/3sGBUzlM+0Mb4iICsZiM+JlLPm3uXMDKtefy9M9Psz5lPX9t/1dubXnr+R5YErA8Z7PZ+OGHH3j33XfZtWsXZ8+eBZCAJYQQ4oou+w2ulPLSWjuAnsAflVIHgVxA4T7JdcXgVbgEzzvAQCAZ2KiUWqi13l1ksy1AR611nlLqQWAG7sWoawZbrrtdg9bwwzPuhaBv+wxM/uc7uuc5XOTbHby8eC/rDp3l6aEx9GwRjq/JSIBPybutnwtYWQVZPP3z0+xJ28OkTpMY1GyQNBktIa01e/bs4fvvvycrK4t27doxYMAALBZZKFsIIcTVXenbdgPQHri5lPvuDBzQWh8CUEp9BowAzocsrfXyItuvA24v5bGqHqcDCrLdP2/9BA4ug+ufhnqt3M+ZA7BjJMdq4/1Vh1i0PYV7e0Yyom1DfLyMBF5DwEq3pjNp1SQOZx7muW7P0btRb2kyWgoul4tly5bh6+vLqFGjaNy4cWWXJIQQohq5UshSAFrrg6Xcd0MublqaDHS5wvb3AouLLUSp+4H7AZo0aVLKciqQ1oWXCbW72eiqGdD8emg3wf26tw8uLwsZuTaW7T3N3DVHuDG+AX/sFYnJaCDQt/SXCM/kneHxVY+TkpPC1B5T6dKgC94Gb4LMQdJk1AMOh4P169fTqVMnTCYTt99+O4GBgRgMMnZCCCFK5krf5nWUUo9e7kWt9etX2Xdx97IXe2eiUup2oCPQ5zLHeh94H6Bjx45V/+7GguwLjUe/+xv4hrrvJlTKvXSOTzBZ+XaOpOYy9bvdxDUIZNLgGExeRoItJe+xdC5gpeSkMHHlRM5az/Jyr5dpV7edNBktgQMHDpCYmEh6ejoBAQHEx8fLcjhCCCFK7Uohywj4U3xY8kQyUPT6SiPgxKUbKaUGAM8AfbTWBaU8VtXhsLnnYgEsexEyj8OoeeAbcn4eVo7NSUaenSe/2o630cArt7TBYjIS7FvygOVwOcgsyORY1jEeX/k4uY5cZvSZQauwVtJk1ENZWVl8//337N69m7CwMCZMmEDz5s0ruywhhBDV3JVCVorWeuo17HsjEKWUisTdZ2ssMK7oBkqpdsB7wGCt9elrOFbV4HK5LxMC7P4Gdn8LXf8EjTu7n/MJokArcqw2Xlm8h0NncnljbFsign1LteCzS7vIKMjgYMZBnlj1BE6Xk1l9ZhEVEiVNRkvgu+++4/Dhw/Tr149u3brh5SU3BgghhLh2V52TVVpaa4dS6s/A97jPis3VWu9SSk0FkrTWC4HXcJ8t+2/hGZxjWuvqu+BlQaa7o3v6YfhpKjTsCF0fdL9msuA0+pCZW8AXm5L5ftcp/q93c7o1DyPY4l3qgLU/fT8TV07E2+DN631fp1lgM2ky6oGjR48SEhJCYGAggwcPRilFSEhIZZclhBCiBrlSyOp/rTvXWicCiZc8N7nIzwOu9RhVhj3f3dndYYPvHnUv9jx0pnsOltEbzIFk5tnZnpzJG0v306NFGHf1aEagrzfeJVzwWWt9/hLhk6ufxGQ08fr1r9PIv5E0Gb2K3NxcfvzxR7Zt20bHjh0ZNmwYoaGhlV2WEEKIGuiyIUtrfbYiC6nWtAZrlvvn1a/BmT0w4u8QUP/89S4lUwAAIABJREFUws/ZBQ5OZ1l56qsd1A00M+WmVviZvUq1XE6WLYtTead4YtUT2F123uj7Bo0DGhNkCsLbWPLWD7WBy+Vi06ZNLFu2DJvNRs+ePenVq1dllyWEEKIGk8knZcGW4+7qfnAZbPkY2t0B1/Vzv+YThNUJ2VYHz327k8w8O/+8syNh/mYCStHNPcuWRVp+GpNWTSLdms7MPjNpHtRcmoxexapVq1i5ciWRkZEMHTqU8PDwyi5JCCFEDSffytfK5XTfTZh9Er5/CurGQa+J7tdMfjiNZrJyC3hv1UE2HknnmWGxxDQIIKgUdxLm2HLIsGbwzJpnOJZ1jGk9p9EqvBUh5hBpMlqM/Px8rFYrISEhdOrUibCwMFq3bi3tLIQQQlQICVnX6lxPrMSJ7t+HvQ5eJjCa0OYAMnJtrNx3hg/XHmVE2whGJEQQ7GvCWMKJ7nn2PLJsWby47kV2pe7i2a7P0rlBZ4LNwRKwLqG1Zvv27fz444+EhoZy99134+fnR5s2bSq7NCGEELWIhKxr4bS7J7xvnQ+/JcHgVyGkWeE8rGCyrA4Op+bywv9207J+AI8NisbfxwuTV8kmulsdVrJsWby28TXWpazjr+3/St/GfQkyBcklwkucPn2axMREjh49SsOGDRkyZIicuRJCCFEp5Bv6WhRkQ24qrJ0DTXtCbGH3Cd9g8h2QkWfjya92YFAw/ZY2BPmYsJhKNuQFzgIyCzL5x7Z/sPTYUu5udTfDrxtOkFkmuV/q0KFDfPrpp5jNZm688Ubat28vAUsIIUSlkZBVWnYrOApg9Uz37/2ecXd0N/tjV95k5Rfw6pK9HDydw+tjEmgSainxmoR2p52sgizm753PV/u/4paoWxgfO55AcyAmo6mcPlj1orUmNzcXf39/mjRpQrdu3ejevTsWi6WySxNCCFHLScgqrYJsOLHZ3dm90x8hJBK8zGiTP5m5Nr7a8huJO05yX89IerQIL/FEd4fLQaYtk4UHFzJ351wGNBnAgwkPSh+sIs6ePcvixYs5c+YMDz30ECaTiQEDak7rNSGEENWbhKzSsOW6z14tewn860OXB8BgLFz42cGO5Exe//FXujYP5Z6ekQT6eONVgoajTpeTjIIMlh9bzpub36RLgy483ulxAk2B0skdcDgcrFmzhtWrV2M0Gunbt68shSOEEKLKkW+mknK5oCAHdnwOp3fDsNlg8gOfIPIcLk5m5fPUVzsI9zczdXhrAn29S9Rw9NxyORtObuDl9S/TKqwVk7tOJtAUiMVbLoFlZ2czb948zp49S+vWrRk0aBABAbIIthBCiKpHQlZJ2XIgLw1+fgMad4XoweDtiw1vMvMKmPztLtJyC/jgjo7UDTTjX4KGo1prMgoy2JW2i+fXPE+TwCZM6zmNYJ9g/E3+5fihqj6Hw4GXl9f5uVfDhg2jefPmlV2WEEIIcVkl6yVQ27mcYM+Dn18Hey70exYMRlymADLz7Xyw+hDrD5/lsUEtad0wiEAfz+/+O7ce4cGMgzy5+klCfEKY3ms6Yb5hBJoCy/FDVW1Op5O1a9fy5ptvkpWVhVKKESNGSMASQghR5cmZrJIoyIKU7bDjC+hwJ4S1AJ9AMq1OVu0/w9w1RxgW34A/tI0gyNcbQwkajmbZsjiec5xJqybhpbyY0XsGDfwa1OqAdfToURITEzl9+jRRUVGVXY4QQghRIhKyPOWwgS0PfpoKfuHQ9c/gZSbH5c2RtGxeWLiLqLr+PHFDS4IsJrxLMNE925bNqbxTTFo1iTxHHrOvn03TwKYEmYNqZZ8nl8vFwoUL2bZtG0FBQYwZM4aWLVvWyrEQQghRfUnI8lRBNuz8Ek7tgMEzwCeAAi9/0rILePLLHbg0TB/ZhhA/U4kmuufZ80jLT+Op1U9xKvcUM/rMoGVoy1oZsLTWKKUwGAwYDAZ69uxJr169MJmkJ5gQQojqR0KWJ+z5kHMaVs+Chh0g9iac3v5kFriY9cOv7DuVzazRCUSG+5doHpbNaeOs9SyT107mQMYBXuzxIgl1Egg2B2NQtWu63IkTJ1i8eDHDhg2jfv363HTTTbUuZAohhKhZJGRdjdbus1hr57jnZPV7Du1lItNp4tstySzcdoK7uzejd3Qdgn09D1jnemG9vP5ltpzewpOdn6R7RHdCzCG1KmBZrVaWLVtGUlISFouFnJwcAAlYQgghqj0JWVdjy4WTO2D7Z5AwDurGkq382JGcwWvf76Nzs1Du7928RBPdtdZk2jKZt2seq39bzYMJD3JDsxsIMgdhNHh+qbG627lzJ0uWLCEvL49OnTrRt29ffHx8KrssIYQQokxIyLoSl8t9FmvZi+ATDN0fwap8OJnt4MkvdxDi583UEa0Isnhj8irBRHd7NquTV/Px7o+5odkNjIoaRZApCG9D7VrwOTU1leDgYMaPH0+DBg0quxwhhBCiTEnIuhJbtnttwhNbYNA0HD7BZDhMTFm4nTPZBbw3oQMNgn2xmDwfxnxHPgczDvLKhleICo7ir+3/SrBPMN7Gmh+wbDYbK1eupGnTpkRHR9OrVy/69OkjlwaFEELUSBKyLsfpcE92XzUTGiSgW91MprYwd+1R1h5M4/EbWtK2cTCBPp4Pod1l50zeGZ5f8zwGZWBK9ymE+YZhMtbsu+e01uzdu5clS5aQlZWF0WgkOjoao7H2XBoVQghR+0jIupyCLFj7lnsJnT+8S5bLh5+PZvHBqkMMbl2fUR0aEmwxeXwWxqVdZFgzmJU0iyNZR5jeazqRQZE1fsHn9PR0Fi9ezP79+6lbty4jR46kSZMmlV2WEEIIUe4kZBXHYXN3dt/yCcSPIT+sNYezDTz37U6uq+PPU0NiCLaYMJako3tBFl/s/4Jlx5dxT+t76BbRDX/vmr8e4dGjRzl69CiDBg2iS5cuGAy1585JIYQQtZuErOJYM2H5S2D2x9n9EdIcZp76eidOl+aVkW0I8zdj9vL8UleOLYekU0m8t+09ukd0Z3zseAJNgTV2LtLBgwfJzc0lPj6ehIQEWrRogb9/zQ+UQgghRFESsi5lz3dPdk/eCANeINe7DrOWHmNPSjYzRsYTVdcff7Pnw1bgLOBY9jGm/jKV+n71mdR5EiHmkBrZqiErK4sffviBXbt20aBBA9q0aYNSSgKWEEKIWklCVlFaQ3YKrJwB9Vphix3Jl3uy+XrLb9zRrSn9YuuWqKO7w+UgLT+Nqb9MJd+Rz2t9XqO+pX6Nu5PQ6XSyYcMGVqxYgcvl4vrrr6dHjx419kydEEII4QkJWUXZcmDtO5B7Goa/xdE8E68v3U3bxsE80Kc5wSVtOFqQyT+2/YNdabt4rutzxITGYPG2lPOHqHgnTpzghx9+ICoqiiFDhhASElLZJQkhhBCVTkLWOS6Xe7L75nnQaiT59drzxqJk8m1OnhwSQ6ifGS+j55O2s2xZLDmyhG8OfMOo6FH0b9KfQFNg+dVfwXJzczl8+DCtW7emcePG3HfffURERMjZKyGEEKKQhKxzrJmw7CXwtqB7PcbK4zYW7Ujhzu5NaRURiI+353Oo8h357E7bzexNs0mok8AD8Q8QZA6qEQFEa83mzZtZunQpDoeDyMhI/Pz8aNiwYWWXJoQQQlQpErLA3Xh097dwbC30fZZ0Uz1e/X4nDYJ8+GOv5gSUYB6W3WXnt+zfeH7t8wSYAni267ME+wTjZaj+Q52SksKiRYv47bffaNasGUOHDsXPz6+yyxJCCCGqpOr/zV8Wck7ByulQJwZn2/H865dUDqfmMmt0AvUCPV+wWGvN2fyzTF03ldT8VGZfP5vGAY0xG83lWHzFyMvLY+7cuZjNZv7whz+cv3NQCCGEEMWTkOUogLVz3HcVDp3JvmwT/1pzmD7RdbihdX28SzgP6x/b/sGW01t4vNPjJNRJwM+7+p7p0Vpz5MgRIiMjsVgsjB49miZNmuDj43nwFEIIIWorab99ahck/QtiR2Br3IPpPxwE4PEbWuJnKtk8rIUHF/Ll/i/5Q4s/MCxyGIHm6jvR/cyZM3z44Yd89NFHHD58GIDo6GgJWEIIIYSHaveZLKcDlk4Bown6PMGifTms+jWVP/dtQcv6AR5fDnO4HGw8uZHZm2bTrm47Hmr7EEHmIAyq+mVYm83GqlWr+OWXXzCZTNx44400a9assssSQgghqp3aHbIOr3T/6vU4aT5NmPHDRpqH+3FPz2YeXybUWnMo4xCT10wmzCeMyV3dv1fHie5aa+bNm0dKSgpt27ZlwIABMrFdCCGEKKXqlwTKitaw9i3wCUZ3uIu3Vh0jJdPKB3d0IMRi8ng3adY0nlv7HLn2XOb0m0PjwMbVrqN7RkYGgYGBGAwGevfujcVioUmTJpVdlhBCCFGtVb/rWWXltyQ4tBw63MX2LAufrDvGsDYN6BNd1+PLhPn2fGZsnMHutN080ekJEuokVKs7CR0OB6tWreKdd94hKSkJgJiYGAlYQgghRBmovWeyfn4DTH7YO9zD1M/2YDEZeXxwS0xenuVOh8vBJ3s+YfHhxYyPHc+QyCHVasmcgwcPkpiYyNmzZ4mLiyMmJqaySxJCCCFqlHI9k6WUGqyU2qeUOqCUerKY13srpTYrpRxKqVHlWctFTu2BfYmQMI4F+zSbjmXw534taBrqWUjSWrMqeRXvbH2Hrg268kDCA/ib/Mu56LKzdOlSPvnkEwDGjx/P6NGjCQysvndCCiGEEFVRuZ3JUkoZgXeAgUAysFEptVBrvbvIZseAu4CJ5VVHsda8AQYvUtvcx6x5v9IqIpA7ujbz+DLh/oz9TF4zmQj/CJ7v+jxBpqByLvjauVwunE4n3t7eXHfddXh7e9OjRw+8vGrvyUwhhBCiPJXnN2xn4IDW+hCAUuozYARwPmRprY8UvuYqxzouln4Mdn4J8bcyfX0Bmfl23p3QAR8Pe2JlWDOYtGoSDu3g5Z4v0zCgYZXvfH7s2DESExOJjIzkhhtuIDIyksjIyMouSwghhKjRyjNkNQSOF3mcDHQpzY6UUvcD9wPXPil7zRugXWxpchdffp7MmE6N6dws1KO32p12Jq+dzMGMg7zc62XahFftpWVyc3NZunQpW7duJTAwkKZNm1Z2SUIIIUStUZ4hq7j0oUuzI631+8D7AB07dizVPgDIOQNb5+OKHc5Tq6yE+5t5/IaWHgUlrTXvbX+P5ceX88c2f2Rws8EYDZ53hK9o+/fv56uvvsJms9GjRw969+6NyeR5awohhBBCXJvyDFnJQOMijxsBJ8rxeFf3y9vgyOdr/7HsPZnNjFHxhPl71nLhp2M/8f729+nbuC8PxD9QZZuNaq1RShEaGkpERASDBw+mTp06lV2WEEIIUeuUZ1LYCEQppSKB34CxwLhyPN6VWbMgaS6O6wYybZORdk0CGNmuoUdvPZF9ghd+eYFmgc2Y2n0qJq+qd0bIarWyfPlycnJyGD16NGFhYUyYMKGyyxJCCCFqrXJr4aC1dgB/Br4H9gCfa613KaWmKqWGAyilOimlkoHRwHtKqV3lVQ8b3oOCLL7wu42zuTYeH9QSowdL5zicDp78+UmsDiuv9HqFYJ/gciuxNLTW7Nixg3feeYcNGzZgsVhwuSruPgIhhBBCFK9cr3lprROBxEuem1zk5424LyOWL7sV1r2LvXEPXt7uT88WwXRvEe7RW9/Z9g5bTm/hqc5P0Sq8VTkXWjKZmZl8++23HD58mIiICG677TYiIiIquywhhBBCUFs6vm/+EPJS+W+TKWRZHUy8Idqjt/1y4hfm7pjLgCYDGNtybDkXWXLe3t5kZmYybNgw2rdvj8FQe1dJEkIIIaqamh+ynHZYOwd7/Xa8uCucQXF1aNs45KpvO5t/lqd/fpoG/g2Y0n1KlQgwWmv27dvHtm3bGD16NBaLhT/96U9VojYhhBBCXKzmh6wd/4XMZD4P/TMFDhcTB7W86lu01jz181NkFmTy7xv+TZC58ju6p6ens2TJEn799Vfq1q1LTk4OgYGBErCEEEKIKqpmhyyt4Zd3sIe0YOqvTRieEEF0/YCrvm3uzrmsPbGWv3X4Gwl1Eyqg0MtzOp2sWbOG1atXo5Ri4MCBdOnSBaOx6vboEkIIIURND1nHfoFTO1nY4DGcWvHYwKufxdp2Zhtvb3mbnhE9ubvV3RVQ5JVprdm+fTvR0dHccMMNspCzEEIIUU3U7JD1yzs4zUE8f6Q1Yzo3pnGY5Yqb59hymLRqEiE+IUzrOa3SlszJzs5m1apVDBw4EJPJxH333YePj0+l1CKEEEKI0qm5ISvjOOxbzLLgkdhzfflL/6grbq61ZvLayaTkpvDBoA8I9fVsPcOy5HK52LBhA8uXL8fpdBITE8N1110nAUsIIYSohmpuyNrwHhrNCyk9uKNXU+oGXjmofPHrF/x49EceSHiAzvU7V1CRFxw/fpxFixZx6tQpWrRowZAhQwgNrfigJ4QQQoiyUTNDli0PNn/ELv8epNrr8ae+La64+f70/by68VU61uvIA/EPVFCRF1uxYgX5+fnceuutxMTEVNqlSiGEEEKUjZoZsrZ/BtZMXsvvzw2t6xNsufxag/mOfB5b+RgWLwszes/AaKiYu/a01mzZsoUWLVoQGBjIiBEj8PHxwWSqeusiCiGEEKLkal7I0hrWvUtGYEtWno7iP52aXHHzV9a/wuHMw/y9/9+pY6lTISWePHmSRYsWkZycTO/evenbt6/cNSiEEELUMDUvZB1eCan7+E/QozQKsdAl8vLzmhIPJfL1ga+5s9Wd9GrUq9xLs1qtLF++nI0bN2KxWLj55puJj48v9+MKIYQQouLVvJC1/n2cPiG8cSqehwY0xmAofm7T8ezjvPDLC7QJb8Nf2v+lQkpbsWIFGzZsoGPHjvTr1w9fX98KOa4QQgghKl7NCln56XDgR7bWuQVbhonRHRsVu5nNaWPiiokYlIHX+ryGt8G73EpKTU3F5XJRt25devfuTXx8PBEREeV2PCGEEEJUDTUrZO1eCE4bb6W1p2dUOBHBxZ8pmrZuGrvP7mb29bNp6N+wXEqx2+2sWrWKtWvX0qxZMyZMmIDFYsFiuXJDVCGEEELUDDUrZO34nDz/pqxIbcg7NxU/4f3zfZ/z1YGvuLvV3QxoOqBcyti3bx+LFy8mMzOThIQEBg4cWC7HEUIIIUTVVXNCVlYKHFnD8pAJBPmaGBBX93ebbDu9jekbptOlfpdym4e1Y8cOvvrqK+rUqcNdd91F06ZNy+U4QgghhKjaak7I2vkloHnzdAI3d26I2eviflep+an8bcXfCPcNZ2afmWXaD8vhcJCRkUF4eDixsbEMGzaMdu3aYTRWTM8tIYQQQlQ9NSdk7fic1IAYfj3TgDc6Nb7oJbvLzqMrHiXTlsnHgz8m2Ce4zA576NAhEhMTsdvtPPzww3h5edGxY8cy278QQgghqqeaEbJSD0DKNr72vZe4iEDiIi5u7Dlj4wy2nN7CSz1eIi48rkwOmZ2dzQ8//MDOnTsJCQnhpptuwsurZgynEEIIIa5dzUgFOz5Ho/hnejse6nVx24aFBxby2d7PGNNyDCNajCiTw6WlpfH+++/jdDrp06cPPXv2lIAlhBBCiItU/2SgNWz/nMP+7Ui3hzOi7YWWDLtTdzN13VTa1mnLpM6TrvlQubm5+Pn5ERoaSufOnWnbti1hYWHXvF8hhBBC1DyGyi7gmp3YAumHSQocwJA2FxaDzrPnMXHVRAJMAcy+fvY1NRzNy8tj4cKFvPXWW2RlZaGUon///hKwhBBCCHFZ1f9M1o4vwGji1gkPMcp8YUL7y+tfJjk7mfcHvk+4JbxUu9Zas2XLFpYuXYrVaqVr1674+PiUVeVCCCGEqMGqf8hK3Qct+oNvyPnTcksOL+Hbg99yZ9yddI3oWqrdOhwOPvroI44fP06TJk0YNmwYdev+vveWEEIIIURxqn/Iuv1LsOWdf5iSk8LUdVOJCY0pVcNRp9OJ0WjEy8uLRo0a0aFDB+Lj41Gq+IWmhRBCCCGKU/3nZAGY3OsBaq15ds2z2J12ZvSegbfxyvOwnDm553/WWrNz507mzJlDSkoKAIMGDSIhIUEClhBCCCFKrPqfySrix6M/suHkBiZ1mkRkUORlt3Pl53Ny2jTyN28h8ov/cjYvj8TERA4fPkyDBg0kVAkhhBDimtWYkFXgLGBm0kyaBzVnbMzYy26Xv3cvJ/72KLbDhwm95x5Wr1vHqjVr8Pb2ZujQoXTo0AGDoWac4BNCCCFE5akxIWveznmk5Kbw/sD38TL8/mNprUn/+GNOz5yFISCAxv/8AP+ePdmzYgWtW7dm4MCB+Pv7V0LlQgghhKiJakTIOpN3hn/t/Be9G/WmW0S3373uSE/nxKQnyV21Ct2nD1t6dMcVHk4M0KdPH7k8KIQQQogyVyNC1uxNs3G4HEzq9Puu7rkbN/Lbo49hz8wk+YEHSMrNQSUn07J1awAJWEIIIYQoF9U+ZO1M3cn/Dv2PO1vdSZPAJhe9lvHVV6RMfp6zMS3Z/IebOZuZQWxsLDfccANBQUGVVLEQQgghaoNqH7J2p+2mnqUeD8Q/cP457XJxZvYbpH3wAZbOnbE+9CB6wwbGjRtHVFRUJVYrhBBCiNpCaa0ru4YS6dixo05KSrroOavDio+Xe7kbZ1YWyc88y9ajRwhISKDv5Mng5YXT6cTLq9pnSiGEELWMUmqT1rpjZdchSq5G9Co4F7Cyl69g/W3j+MZoYGuHDqTFx6O8vVFKScASQgghRIWqEcnDmZnJkWnTWJNyksOdO+Hv68voG28kNja2sksTQgghRC1V7UOWdc8ejv3xfk4ajRzt15duXbrQp29fzGZzZZcmhBBCiFqsXEOWUmow8CZgBP6ptZ5+yetm4COgA5AGjNFaH/F0/8e3bWfbKy/T0mik8z/+TttGjQgMDCy7DyCEEEIIUUrlFrKUUkbgHWAgkAxsVEot1FrvLrLZvUC61rqFUmos8Cow5mr7tlqt/DBvHltTUjBHRtJ78mR84uLwKY8PIoQQQghRCuU58b0zcEBrfUhrbQM+A0Zcss0I4MPCn78A+qurdAfNy8hgzksvseXkSaJS0/jj6FsJjIsr8+KFEEIIIa5FeV4ubAgcL/I4GehyuW201g6lVCYQBqRebqeZeXn4WK3cFB9PzHPPoYzGMi5bCCGEEOLalWfIKu6M1KVNuTzZBqXU/cD9hQ8L/jJr1k4A7rzzWuqrbcK5QngVlyXjVnIyZqUj41Y6tWHcmlZ2AaJ0yjNkJQONizxuBJy4zDbJSikvIAg4e+mOtNbvA+8DKKWSpClbycm4lY6MW8nJmJWOjFvpyLiJqqw852RtBKKUUpFKKRMwFlh4yTYLgXOno0YBy3R1a0EvhBBCCFGMcjuTVTjH6s/A97hbOMzVWu9SSk0FkrTWC4F/AR8rpQ7gPoM1trzqEUIIIYSoSOXaJ0trnQgkXvLc5CI/W4HRJdzt+2VQWm0k41Y6Mm4lJ2NWOjJupSPjJqqsardAtBBCCCFEdVAjFogWQgghhKhqJGQJIYQQQpSDKhuylFKDlVL7lFIHlFJPFvO6WSm1oPD19UqpZhVfZdXiwZj1VkptVko5lFKjKqPGqsiDcXtUKbVbKbVdKfWTUkp61uDRuD2glNqhlNqqlPpZKSVLM3D1cSuy3SillFZK1fr2BB78WbtLKXWm8M/aVqXUfZVRpxCXqpIhq8i6h0OAOOC2Yv6CPr/uITAb97qHtZaHY3YMuAuYX7HVVV0ejtsWoKPWOh738k8zKrbKqsfDcZuvtW6jtW6Le8xer+AyqxwPxw2lVADwCLC+YiusejwdM2CB1rpt4a9/VmiRQlxGlQxZlNO6hzXcVcdMa31Ea70dcFVGgVWUJ+O2XGudV/hwHe7GurWdJ+OWVeShH8Ws5lALefJ3G8CLuIOptSKLq6I8HTMhqpyqGrKKW/ew4eW20Vo7gHPrHtZWnoyZ+L2Sjtu9wOJyrah68GjclFJ/UkodxB0YHqmg2qqyq46bUqod0Fhr/V1FFlaFefr/6MjCS/pfKKUaF/O6EBWuqoasMlv3sBaR8Sgdj8dNKXU70BF4rVwrqh48Gjet9Tta6+uAScCz5V5V1XfFcVNKGXBPf3iswiqq+jz5s/Y/oFnhJf2lXLjKIUSlqqohqyTrHnKldQ9rEU/GTPyeR+OmlBoAPAMM11oXVFBtVVlJ/7x9BtxcrhVVD1cbtwCgNbBCKXUE6AosrOWT36/6Z01rnVbk/8sPgA4VVJsQV1RVQ5ase1hynoyZ+L2rjlvh5Zv3cAes05VQY1XkybhFFXk4DNhfgfVVVVccN611ptY6XGvdTGvdDPccwOFa66TKKbdK8OTPWoMiD4cDeyqwPiEuq1yX1SktWfew5DwZM6VUJ+BrIAS4SSn1gta6VSWWXek8/LP2GuAP/Lfw3opjWuvhlVZ0FeDhuP258AygHUjnwj+Kai0Px00U4eGYPaKUGg44cH8f3FVpBQtRhCyrI4QQQghRDqrq5UIhhBBCiGpNQpYQQgghRDmQkCWEEEIIUQ4kZAkhhBBClAMJWUIIIYQQ5UBClhAVTCnlVEptLfKr2RW2baaU2lkGx1yhlNqnlNqmlFqjlGpZin08oJS6o/Dnu5RSEUVe++dlFu29ljo3KqXaevCevyqlLNd6bCGEKGsSsoSoePla67ZFfh2poOOO11on4F5ypMRLA2mt39Vaf1T48C4goshr92mtd5dJlRfq/Due1flXQEKWEKLKkZAlRBVQeMZqtVJqc+Gv7sVs00optaHw7Nf2cx3VlVK3F3n+PaWU8SqHWwW0KHxvf6XUFqXUDqXUXKWUufD56Uqp3YXHmVn43BSl1ESl1Cjcazh+WnhM38IzUB2VUg8qpWYUqfkupdRbpazzF4osBKyU+odSKkkptUsp9ULhc4/gDns1AbLIAAADF0lEQVTLlVLLC58bpJT6pXAc/6uU8r/KcYQQolxIyBKi4vkWuVT4deFzp4GBWuv2wBhgTjHvewB4U2vdFnfISVZKxRZu36PweScw/irHvwnYoZTyAeYBY7TWbXCvAPGgUioU+APQqnDB3ZeKvllr/QWQhPuMU1utdX6Rl78AbinyeAywoJR1Dga+KfL4Ga11RyAe6KOUitdaz8G9jl1frXVfpVQ47oWoBxSOZRLw6FWOI4QQ5aJKLqsjRA2XXxg0ivIG3i6cg+QEoot53y/AM0qpRsBXWuv9Sqn+uBfD3Vi45I8v7sBWnE+VUvnAEeBhoCVwWGv9a+HrHwJ/At4GrMA/lVKLgO88/WBa6zNKqUNKqa641ypsCawp3G9J6vTDvYRK+yLP36qUuh/331sNgDhg+yXv7Vr4/JrC45hwj5sQQlQ4CVlCVA1/A04BCbjPMFsv3UBrPV8ptR73YsvfK6XuAxTwodb6KQ+OMb7oQsNKqbDiNipcK64z0B/3mqB/BvqV4LMsAG4F9gJfa621cicej+sEtgHTgXeAW5RSkcBEoJPWOl0pNQ/wKea9CvhRa31bCeoVQohyIZcLhagagoAUrbULmID7LM5FlFLNgUOFl8gW4r5s9hMwSilVt3CbUKVUUw+PuRdoppRqUfh4ArCycA5TkNY6Efek8uLu8MsGAi6z36+Am4HbcAcuSlqn1tqO+7Jf18JLjYFALpCplKoHDLlMLeuAHuc+k1LKopQq7qygEEKUOwlZQlQNfwfuVEqtw32pMLeYbcYAO5VSW4EY4KPCO/qeBX5QSm3///btEAehGIii6O1KWBt7YBt4giA4BJYEgSA4giGEwCYQWBSDaL/5hiAGEnKPrJg2VS/TKbChPqW9FREPYAgsSyln4AlMqIFl1ertqF22vjkw6Qbfe3XvwBUYRMShrX18zjbrNQZGEXECjsAFmFGfIDtTYF1K2UbEjfrzcdH22VPvSpK+rkTEr88gSZL0d+xkSZIkJTBkSZIkJTBkSZIkJTBkSZIkJTBkSZIkJTBkSZIkJTBkSZIkJXgB5u0o77MCyQIAAAAASUVORK5CYII=\n", + "image/png": "iVBORw0KGgoAAAANSUhEUgAAAkYAAAFACAYAAABZbLPsAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4wLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvFvnyVgAAIABJREFUeJzs3Xd4lFX2wPHvnZpegABCqKEFAhEIIF0QaQoqIIu4lt117WVt69pX197Wvi6Wdf1ts9fQUUE6EaQTeg2d9Eyf+/vjjZjJFEIJk8D5PA/Pkpn7vnNZk5mTe889R2mtEUIIIYQQYIr2BIQQQggh6goJjIQQQgghKklgJIQQQghRSQIjIYQQQohKEhgJIYQQQlSSwEgIIYQQopIERkIIIYQQlSQwEkIIIYSoJIGREEIIIUQlS7QncLwaNWqkW7duHe1pCCGizKc1fj9YzSraUxHilPP5fJSUlOBwONi7d+8hrXVatOd0tqh3gVHr1q3Jy8uL9jSEEFHi9PgodXrxV7YzSkuwYzJJcCTODD6fj8WLFzN37ly01gwaNIhBgwbtiPa8zib1LjASQpydfH5NqdODy+uP9lSEqDUfffQRGzdupGPHjowcOZKUlJRoT+msI4GREKLOK3d5KXd5qd7y2ufXHChx0jQlNirzEuJUKCsrw263Y7Va6dOnDz169KBjx47RntZZSwIjIUSd5fH5KXF48Pqrh0Sw7VA5T+aux+Pz8/VtA7Ca5SyJqF/8fj95eXl8++239O7dm6FDh9K2bdtoT+usJ4GREKLO0VpT6vLicPuCnvP6/Pzf4h28O38bsTYzf764CxbJMRL1zJ49e8jNzWXv3r20adOGbt26RXtKopIERkKIOqV6cnVV6/eW8ETuejYfKGNYZmPuHt6Rjk0SUUoCI1F/LF68mBkzZpCQkMD48ePp0qWLfA/XIRIYCSHqhEjJ1U6Pj3fnb+Pfi3eSGm/lufHdGNxRTi+L+kNrjcfjwWaz0aZNG/r06cOQIUOw2+3RnpqoRgIjIUTUhUuuBlixs5Anp65n1xEHl5zbjNuGtiMxxnra5yjEiTpw4AC5ubkkJyczbtw4mjRpwsiRI6M9LRGGBEZCiKhxe/2UOkMnV5e5vLz53WY+Xb6HZikxvH5Fd3q1aRCFWQpxYlwuF3PnzmXx4sXExMSQnZ0d7SmJGpDASAhx2kVKrgZYsPkQz07fwIESF1f0bsENgzKItZkDxiggIcYixR1FnbR7924++ugjSktL6dGjBxdccAFxcXHRnpaoAQmMhBCnVaTk6qIKN3+dvYnpa/bRplE8b1/Tla7Nk4PGWc0mkmIsWOSIvqhjtNYopUhJSaFBgwZMnDiR9PT0aE9LHAcJjIQQp0Wk5GqtNXPWH+CFmfmUOL38bkAbru3XGpslMPBRQLzdQrxd3rpE3eLxeJg/fz67du3iqquuIiEhgWuvvTba0xInQN5dhDib+Lzg94I15rS+bKTk6oOlLp6bsYF5Gw+ReU4ir03OpH3jxKBxskok6qpNmzYxbdo0CgsLycrKOnr6TNRPEhgJcbZwV4CrBGzxwOkJjCIlV2ut+XrlXl6ZswmPz89tQ9sxqXcLLCZZJRL1Q0VFBd988w3r16+nYcOGXH311bRp0yba0xInSd5phDjTaQ3OYvA4TuNLRk6u3lPo4Kmp68nbUUiPlik8MDqTFg2CE1MtJkVyrFVWiUSdZLVaOXToEEOHDqVfv36YzeZjXyTqPAmMhDiT+bzgKDS2z06TSMnVPr/mo7xd/O37LZhNij+N6sQl5zbDVK3qrwLi7BYSZJVI1DE7duxg4cKFTJgwAavVyo033ojJJIH7mUTedYQ4U3kcxkpRiAClNkRKrgbYerCMJ3LXs7aghP7tGnLfyE40SQre0rOYFEmxVmkKK+qUsrIyZs2axapVq0hOTqaoqIi0tDQJis5AEhgJcaaJwtZZpORqj8/PPxdu5x8LtpNgt/D4JV0Y3rlJyN5Q8XYL8Taz9I0SdYbWmry8PObMmYPH42HgwIEMHDgQq1Wqr5+pJDAS4kzi84KzCHye0/JykZKrAdYVlPBk7no2HyxjeOcm3HVhB1Ljg0/rmCtziWSVSNRFq1atolmzZowePZpGjRpFezqilklgJMSZoiZbZ4c3w9xnYMI/IKHxCb/UsZKrnR4fU+Zt5b9Ld9Iwwc4Ll3djYPvQTV/jbGYS7BZZJRJ1hsPhYO7cuQwYMICEhAQmT55MTEyMfI+eJSQwEqK+q+nW2YZcmPUwWOOgeNcJB0aRkqsBftxRyFNT17O70MFl3Ztz65B2JMQEv9WYTYqkGGtQEUchokVrzapVq5g5cyYOh4PmzZvTtWtXYmNjoz01cRpJYCREfVaTrTOfG+Y+Bz/9C5p1h3FToFGH43+pYyRXlzm9vPbtJr74qYD01FjemNydnNahm77G2swkyiqRqEMOHDhAbm4uO3fuJD09nYsuuoimTZtGe1oiCiQwEqK+qsnWWek++OYO2LsSelwDA++BuNTjfqlIydUA8zcd4pnpGzhc5uLKPi25flBbYqzBNV1MysglklUiUdfMnz+fgwcPMmbMGLp37y5B+1lMAiMh6puabp3tWAhT7wavCy5+GTqMPO6XOlZydWG5m5dmbWTmuv1kpMXz7PiudGkW3PQVZJVI1C1aa9avX0+jRo1o3LgxI0aMYOTIkcTFBRcaFWcXCYyEqE9qsnWm/bB0Cix4BRpmwJhXoUHb43oZv19T5g6fXK21Zua6/bw0cyNlLi+/H9iGa/q1DnmqzKQUSbEW7BapCizqhiNHjjBt2jQ2b95Mjx49GDNmDPHx8dGelqgjJDASor6oydaZowim3wfb5kKni2HYY5W90WruWMnV+0ucPDc9n/mbD9GlWRIPjs4ko3FCyLExVjNJMbJKJOoGr9fL/PnzmT9/PmazmZEjR9KrV69oT0vUMRIYCVHXaW00f3VXRB63fw18fQeUHYChj0D2FXAcAcmxkqv9WvPlTwW89u0mvD7NH4a1Z2JOC8ym4NeQVSJRFy1ZsoS5c+eSlZXF8OHDSUxMjPaURB0kgZEQdVmNts40rP4YvnsC4hrCr/4F52Qf18scK7l615EKnpq6nuU7i8hplcoDozNpnhr6CHOM1cglMoUImIQ43UpKSigrK6NZs2b07t2bZs2a0aZNm2hPS9RhEhgJUVfVZOvM44A5j8G6L6BVfxj9AsRGOHVmthl1jCodK7na6/fzv6W7mDJvKxaz4oHRnRib3Szk1phSkBRjDXkaTYjTzefzsWTJEr7//ntSU1O58cYbsVqtEhSJY5LASIi6pqZbZ4U74Ovb4VA+nHcznHcLmCIEJbY4sCeBUsesXA2w+UAZT+auZ93eEga2b8QfR3akcWJw01eAGIuZxBhZJRJ1w44dO8jNzeXgwYN06NCBkSNHSp6bqDEJjISoS2ra62zzbJj+JyMQumwKtBkUfqxSRkBkM1aKfH5NscODxxc6l8jt9fP+wu28v3A7STEWnrg0i2GZjWWVSNQL27Zt44MPPiA5OZlJkybRsWPHaE9J1DO1GhgppUYCrwBm4B2t9TPVnr8WeB7YU/nQ61rrd2pzTkLUWTXZOvN7Yf7LkPcONMky6hMlp4cfbzIbW2tmoxO42+unyOEO+xJr9hTzZO56th4qZ2RWU+4c1p6UuOCmrwB2i4mkGKusEomo8/v9HD58mLS0NFq3bs2oUaM499xzsdlCf+8KEUmtBUZKKTPwBnAhsBtYppT6Smu9rtrQD7XWt9bWPISo82q6dVZ+EHLvht1LodskOP8BsER447fYISYFTEZtoXKXlzKXN+RQh9vH3+dt4X9Ld5GWaOelidn0bxe6i7isEom6pKCggNzcXI4cOcJtt91GXFwcvXv3jva0RD1WmytGvYHNWuutAEqp/wGXANUDIyHOXn4fOAqPvXW2Ow9y7wRXKYx8BjpfGnm8LR5ikgCjGGOJw4vTGzqfKG/7EZ6auoE9RQ7G92jOzUPakWAP/dYgq0SirnA4HHz77bfk5eWRkJDA6NGjpdmrOCVqMzBqDuyq8vVuoE+IceOVUoOAjcCdWutd1Qcopa4Hrgdo2bJlLUxViCjwOCu3zkLn+gDGatLy92HeC8aW2bh3IC1CzoRSEJMMVuMDwuvzU+Tw4Atx6qzU6eG1bzfz5U8FtGgQy1u/7kH3lqFPtCkFiXYrsTZZJRLRV15ezt/+9jcqKiro3bs3Q4YMISYm9MEAIY5XbQZGoX6lrP7u/DXwX621Syl1I/BPYGjQRVpPAaYA5OTkREjAEKIeqOnWmasMZj4Am2ZCuwthxFNgj1CQzmSpzCcyfqydHh8lTk/IfKIfdxTy6JdrOVzu4qrzWnHdwDZht8ZsZhNJsdaQhRyFOJ0qKiqIi4sjPj6eXr160aFDB84555xoT0ucYWozMNoNtKjydTpQUHWA1vpwlS/fBp6txfkIEX013To7mG8cxS/eDYP+CD1/E7mKdbV8orLKgo3Vaa351+KdvPn9ZlqkxvH85b3IPCcp5C0VkBgjq0Qi+txuN3PnzmXp0qX8/ve/p3HjxgwePDja0xJnqNoMjJYB7ZVSbTBOnU0CJlcdoJQ6R2u9t/LLscD6WpyPENFVk60zgHVfwexHjNWhy9+H9GP0crInHF1J8lcexXeHOIpf5vTy+DfrmLvxIBd0asyDF2USHyaXSFaJRF2gtWbDhg1Mnz6dkpISunfvTkJC6L58QpwqtRYYaa29SqlbgRkYx/Xf01qvVUo9DuRprb8CbldKjQW8wBHg2tqajxBRU9OtM68b5j4NK/8LzXPgopcgoXH48UoZq0RWI7fC4/NTVOEJ2fx104FS/vTpavYWO/nDsPZM6tUidF0iICHGQpxNSpyJ6NJa8+GHH5Kfn0+TJk2YMGECLVq0OPaFQpwkpSPVTKmDcnJydF5eXrSnIUTN1HTrrGQPfP0H2L8acn4HA+40cobCqZZP5HD7KHV6QvY6y121l2enbyApxsqTl2WR3SIl5C2tZhNJMRYsZlMN/3FCnHo+nw+z2di+nTdvHjabjd69e2Mynb3fl0qpH7XWOdGex9lCfi0UorbUdOts2w8w7R4jiBrzGrS/MPJ4i90Iio7R2sPl9fHSzI188VMBPVul8pdLutAwwR40TlaJRF2xefNmpk2bxujRo8nIyGDQoAgV3YWoJfJOKMSpVtOtM78PFr9p/GnUHsa8CqmtI19jTzRyiojc2qOgyMH9n61mw75Sru7bihsGt8US4jduWSUSdUFJSQkzZsxg3bp1NGjQAItFPppE9Mh3nxCnkt8HjiLwuSOPcxTC1Hthx3zofAlc8OejtYdCUiaITTFWizBWg4odoY/iL9xyiEe/XItfw3MTujG4Q1rIWybYLWGTr4U4XfLy8pg5cyZaa4YMGUK/fv0kMBJRJd99QpwqXrcR8Bxr62zvKvjmD1BxEIY9Bl0nRj6Kb7YaW2cmI+8iXGsPn1/z7vxtvDd/GxmNE3hmXFdaNIgLGmdSiuRYKzaLrBKJuuHn/mapqaELjApxOklgJMSp4C432nVEOsygNaz6H3z3FCSkwa/+A027Rr6vNcY4eVaZTxSutUdRhZtHvlzLkm1HuKjbOfxxRMeQBRttZhPJsdLSQ0RPeXk5s2fPJj09nZ49ex79E+qUpBDRIIGRECejpvlEngqY/WdY/xW0HgSjnjVWgSKJSTJ6nhG5tcfagmLu/2w1heUeHhjdibHZzUJ+yMTbLWF7oAlR2/x+P8uXL2fOnDm43W4aNmwIIAGRqHPkXVKIE+X3Vx7FP0Y+0ZGt8PUdcHgz9Lsd+txo5AyFo0xG0GSxAZWtPRzBR/G11ny6fA9/nbWRtEQ7U67uGbKKtVKQHGvFbpEK1iI69u3bx9dff01BQQGtW7dm9OjRpKWFzn0TItokMBLiRPg8RlDkD92x/qiNM4x+ZyYrjH8HWvWPPN5sM5KsK/OJSp0eKkIcxXe4fTwzbQPT1+6jX0ZD/jy2C8mx1qBx1sqtM6lgLaKpvLyc4uJixo0bR1ZWlqwSiTpNAiMhjpfHUVmfKEI+kc8D81+EH9+Hptlw8V8hqVnk+1pjISYZlIrY2mPn4Qru+3QV2w6Vc8OgtlzbvzWmEB80cTYziTHBwZIQtU1rzerVqyktLaV///5kZGRwxx13YLXK96Oo+yQwEuJ4uEqNrveRlO2H3Ltgz49w7pUw+D5jJSgcpcCeBDbjBFmk1h7fbjjAX75Zh9Vs4pUrzqVPm4Yhb5cUYw2ZfC1EbTt48CC5ubns2LGDli1b0rdvX0wmkwRFot6QwEiImtDa2DrzuiKP27XECIrcFTDqecgcE3m8yWycOqvMJwrX2sPr8/PG91v4z5KddGmWxNPjutIkKSbodhaTIiXOJltn4rRzu93MmzePRYsWYbPZuPjii+nRo4dsm4l6RwIjIY7F563MJwquHXSU1rD8nzDveUhpCRPeN6pZR2K2VdYnMkVs7XGozMWDn6/hp11FXN4znTuGtccaolJ1rM1Mot0iH0QiKoqLi1m0aBHdunVj2LBhxMfHR3tKQpwQCYyEiMTrMipZRyra6HHArIdhwzeQMQxGPnO0bUdYtjgjn4jIrT2W7yjkwS/WUOH28vglXRjRpWnQGAUkxcrWmTj9CgsLWb9+Pf369SMtLY3bb7+d5OTkaE9LiJMigZEQ4bjKjJyiSIp3w1e3wcEN0P8P0Pv6YxzFV0ZAVNn+I1xrD601/1qyk799t4XmqbG8Mbk7bdOCgy2zSZESa5VeZ+K08nq9LFiwgPnz52MymcjKyiIpKUmCInFGkMBIiOq0Nk6deRyRx+1YaOQTaR9c+ha0HRx5vMlsbJ2ZjSTUcK09ypxe/vLNOr7feJChnRrz4EWZIQszxljMJMXK1pk4vbZs2cLUqVM5cuQIXbp0Yfjw4SQlBdfPEqK+ksBIiKpq0gRWa/jxH/DDC9CgLYx9HVJbR76vxW4kWVfmE4Vr7bHpQCl/+nQ1e4uc/GFYeyb1ahEU+CggMcZKrE22zsTp5XK5+OSTT4iLi+PXv/41GRkZ0Z6SEKecBEZC/MzrBmdR5KKNHgfMfAjyc6H9cBjxFNiOlU8Ub7T3IHJrj6mr9/LMtA0kxlh489c9OLdFStAYs8loABsq+VqI2uDz+VizZg3dunXDbrdz1VVX0bhxYywW+fgQZyb5zhYCala0sXg3fHUrHMyHAXdCr+uNnKFwlDJWiazGsfpwrT3cXj9/nbWRz1bsoUfLFJ64NIuGCfag29ktRhVr2ToTp8vOnTvJzc3lwIEDxMXF0b59e5o1O0ahUiHqOQmMhHCWgLs88pgdCyrziTRcNgXaDIw83mSpzCcyfsTCtfbYW+zg/s9Ws35vKVf3bcUNg9tiMQWuBikgIcZCnE1+XMXpUV5ezuzZs/npp59ISkriV7/6Fe3atYv2tIQ4LeSdVpy9/H5j6yxS0UatIe9dmP8SNMiozCdqFfm+FrsRFCmF1sZRfJc3+Cj+oi2HeeSrNfj8mufGd2Nwx+CmmiZlbJ3ZLLJ1Jk4PrTX/+c9/2LdvH/3792fQoEHYbBEqtwtxhpHASJydalK00VMBMx6EjdOgw0gY/qSRLxSJPQHsicZL+DWFFe6gfCK/1rw3fxvv/LCNjMYJPD2uKy0bxAXdylbZANYkVazFabB3714aNWqE1Wpl5MiRxMTEkJYWHKwLcaaTwEicfTxOY6UoUj5R0U4jn+jwZhh4D+T87hj5RCaITTFWizDyhooc7qCXKK7w8OhXa1m09TCjuzblvpGdQhZmjLdbQh7RF+JUczqdfPvtt+Tl5TF48GAGDx5MixYtoj0tIaJG3nnF2aUmRRu3/wC59xh/v2wKtB4QebzZWtnawwhwwvU7W1dQwv2freZwuYs/jerEpec2Cz6KryA51ordIkfxRe3SWrN69WpmzpxJRUUFOTk59OnTJ9rTEiLqJDASZwetjVUijzPymGVvw/y/Gn3Oxr4BKcf4zdkaa1SyrgxwSpyeoH5nWms+X7GHl2ZtpFGCnbevziHznOCCeNbKrTNpACtOh1mzZrFo0SKaNWvG5MmT5bSZEJUkMBJnPr/PyCfyecKPcZfDjAdg0wzoOBqGPwHW4Lyfo5QCe5LR8wwj+Cmq8OCu1u/M5fXx7LR8clfvpW/bhjw2tgvJcdag28XZzCTGBD8uxKnkdrvxer3ExcVx7rnn0qBBA3r06IHJJMn9QvxMAiNxZvO6jaAoUhPYwh1GPtGRLTDoXuj528j5RNVae4RLsj5S7ua+T1exancx1w1ow+8GtsEUYussKUYawIrapbUmPz+fadOm0bJlS8aPH0/jxo1p3LhxtKcmRJ0jgZE4c7krwFUSOcl62zyYeo8RoYx7G1r1j3zPKq09IHyS9eYDZdzz8UqOlLt56rIsLshsEnQr2ToTp0NhYSHTpk1j06ZNNG7cmF69ekV7SkLUaRIYiTOTs9gIjMLRGpb+HRa8AmkdjfpEyemR71nlKD6ET7Kev+kQD3+5hnibhb9f1TNkPlGszUyiXRrAitqVn5/PJ598gslkYvjw4fTu3RuzWVYnhYhEAiNxZqlJ0UZ3GUy/HzbPgo4XVeYTxYYfr0xGgnVlaw8In2T9n6U7eW3OZjo2TeT5y7vRODEm8FZAUqxsnYna5fF4sFqtNG/enKysLIYMGUJSUnCALoQIJoGROHP4PJVFGyM0gS3cXplPtBUG3wc9ro2cT2S2Gltnla09/H6jknX1JGuPz89z0/P5amUBQzs15tExnYOCH0tlA1iLNIAVtaSkpISZM2dSUlLCb37zGxISErjkkkuiPS0h6hUJjMSZoSZFG7d+D9PuBWWG8e9By/Mi39MaYwRFlYGT1+enyOEJSrIurvDwp89WsXxnEb/p35rrB7UNSrKOsZpJipGtM1E7/H4/S5Ys4fvvv8fn8zFw4ED8fr9smwlxAiQwEvWfq9Qo3BiO9sOSt2Dha5DWCca+FjmfSCkjl6hK+w+X10exwxMUd207VM49H6/kQImLxy/pwoguTYNulygNYEUtKiws5MMPP2T//v20a9eOUaNG0aBBg2hPS4h6S96tRf3mKIxctNFVBtP/BFtmQ6cxcOHjkfOJTGZjlcjyS9PMcEnWi7ce5oHPV2O3mHnz1z3o2jw54HmpYi1qk9YapRQJCQnExsYyceJEOnXqJKuSQpwkCYxE/eU4RiXrI1uNfKLCHXD+A9D9qsj5RNWO4kPoJGuAj/N28dKsjbRNS+DFy7NpmhyYZG02KVIkn0jUAq01y5cv58cff+Q3v/kNVquVa665JtrTEuKMIYGRqJ+cxeBxhH9+y7cw7Y9G8vT4d4+dT2SLh5hfTu2ES7L2+vy8NGsjny7fw8D2jXj8ki5B22RWs4mUWCsmqU8kTrG9e/eSm5vLnj17aNWqFQ6HA6tVKqYLcSpJYCTqH1dp+BpF2g+L34RFr0PjzkZ9oqQIPaCUMlaJqhzFD5dkXeLw8ODna1i6/QhXndeKm87PCCrOKEnWojZ4vV5mzZrFsmXLiIuL47LLLqNr167yfSZELZDASNQvrrLwidauUph+n7Fa1PkSuOCxgIAniMlS2drjlx+DcEnWO49UcPdHKykocvDwxZlc3C042EqwW4i3y4+UOPXMZjP79+8nJyeHoUOHEhMT4ftaCHFS5F1c1B/uciP4CeXIVvjyFijaCUMehHN/HTmfqNpRfIAKt5cypzcoyTpv+xHu/2w1Silen9yd7i1TA56Xoo2iNhw8eJA5c+Zw8cUXk5CQwFVXXSXH74U4DWo1MFJKjQReAczAO1rrZ8KMmwB8DPTSWufV5pxEPeWuAGdJ6Oc2z4HpfwSzHSb8A1r0jnwve6LR3qOKcEnWn6/Yw/Mz8mnZII4XL8+meWrgiTaTUqTEWbFKkrU4RTweD/PmzWPhwoXYbDYOHDhAQkKCBEVCnCa1FhgppczAG8CFwG5gmVLqK631umrjEoHbgSW1NRdRz3kcRrJ1ddpv5BItfhOaZBn1iRLPCX8fZTK2zqocxQ+XZO3za16Zs4kPl+2ib9uGPHFpFgkxgT8uFpMiJc4mTWDFKZOfn8+0adMoLi4mOzubCy+8kPj4+GNfKIQ4ZWpzxag3sFlrvRVAKfU/4BJgXbVxfwGeA+6pxbmI+srjDBMUafj2CVj5H+hyGVzwZ+O4fThmG8SmGHWKKoVLsi5zeXnoizUs2nKYSb1acNsF7bCYAleEYixmkmIlyVqcWmvXrsVut3PttdfSqlWraE9HiLNSbQZGzYFdVb7eDfSpOkAp1R1oobX+RiklgZEI5HWFbvOhNcx9xgiKev4WBt0bOZ/IFgf2pIAx4ZKs9xQ6uPvjlew8UsF9Izsyrkdwhex4u4UESbIWp4DX62XhwoV07NiRJk2aMHr0aKxWq2ybCRFFtfnuHuqT6ujHkFLKBPwVuPaYN1LqeuB6gJYtW56i6Yk6zes2qlqHCormvwTL/2kUbIwUFCkFMclBla4r3F5Knd6g4St2FvKnT1fj15pXJ51LTuvAtgqSZC1Opa1btzJ16lQOHz6M1pomTZrIaTMh6oDaDIx2Ay2qfJ0OFFT5OhHIAr6v3I5oCnyllBpbPQFbaz0FmAKQk5MToUuoOCP4PKGDIoBFr8Gyt6HbJKOadbigyGQxts7MgcXvwiVZf7OqgKenbqBZSiwvXp5Ny4ZxAc8rBSmxNmwWSbIWJ6e0tJQZM2awdu1aGjRowJVXXkm7du2iPS0hRKXaDIyWAe2VUm2APcAkYPLPT2qti4FGP3+tlPoeuEdOpZ3lfF6oOGIkVle35C0j0brLeLjgkfBBUYjWHpGSrP/2/Rb+b/EOerVO5anLupIUGxhMmU2KVEmyFqfIjz/+yIYNGzj//PPp378/FotsywpRl9TaT6TW2quUuhWYgXFc/z2t9Vql1ONAntb6q9p6bVFP+X3gCBMULXsXFrwMmWONRrAqzMqNPcE4jl9FuCTrCreXR75cyw+bDjGue3PuHt4hqLeZzWwiJc4qSdbipOzatQu/30+rVq3o378/3bp1o0GDBse+UAhx2tXqrypa66nA1GqPPRJm7Pm1ORdRx/l9xkqRP3ibi+UfwA+eNcjsAAAgAElEQVTPQ4dRMOKpgJNlRymTsXVW7WRauCTrfcVO7vl4JVsOlnHP8A5M6JkeFPzE2swkxUgfKnHiKioqmD17NitWrKBVq1Zce+21WK1WCYqEqMNkDVdEn99fGRQFJ0Sz8r/w/VPQ7kIY9ZyRO1Sd2WrUJ6oWMIVLsl69p5g/frIKl9fHX391Lue1bRg0JjHGEtQcVoia0lqzfPly5syZg8vlol+/fgwePDja0xJC1IC884vo8vuN7bNQQdHqT2DOY9D2fLjoxaBEasA4cRaTHJRvFC7JesbafTzxzXrSEu28Mbk7bdMCK2ArBcmxVuwWOXkmTtyGDRv45ptvaNWqFaNHj6Zx48bRnpIQooYkMBLRo7Vx+sznCX5u3Zcw62FoNQAufsUo0FidNdbYPqsiXJK1X2venreV9xZsp3uLFJ4Z35WUuMB7mpQiNc4alGckRE04nU72799Pq1at6NSpE5MmTaJDhw6SnyZEPSOBkYiOo0GRO/i5/Kkw435o0QfGvh66onWIoChckrXT4+Oxr9fx7YYDjMk+h/tGdgrqbWY1m0iJtWKSk2fiOGmtWbNmDTNnzsTn8/GHP/wBm81Gx44doz01IcQJqFFgpJSyAS211ptreT7ibOEoNCpbV7dpJky9F5r1gEvfBGuIgnchgqJwSdYHSp3c+/Eq8veVcvsF7Zjcu2XQb/AxVjNJMdLeQxy/Q4cOMXXqVLZt20azZs246KKLsNlCrG4KIeqNYwZGSqmLgJcAG9BGKXUu8KjW+rLanpw4Q4ULirZ8C7l3Q9OucNlbYI0LHhMiKAqXZL1+bwn3fryKcreXFy7PZkD7RkFjEuwW4qW9hzgBRUVFvPXWW1itVkaPHk3Pnj0xmWQbVoj6riafCI9j9Dj7DkBr/ZNSSsq0ihPjKDIaw1a37Qf45g5I6wjj3gZbQvCYEEFRuCTrOev389jX60iNszHl6p60bxxY20jae4gTdfDgQdLS0khJSWHkyJFkZmYSHx8f7WkJIU6RmgRGHq11UbVtBmnLIY6fswQ8juDHdy6Cr26FBu1g/LtBBRqBoKAoXJK11pp/LNjO3+dtpWvzZJ4d35WGCYE5SialSImzBuUZCRFJYWEh06dPZ9OmTdx0002kpaWRk5MT7WkJIU6xmgRG65VSEwFTZXuPO4DFtTstccZxlYK7PPjx3cvgi5sgpSVMeM84el9dtaAoXJK1y+vjydz1zFi7n5FdmvLARZ2Cjt1bTIqUk2jv4dd+SlwlJNmTMIWrvi3OKF6vl0WLFjFv3jyUUgwbNkwKNApxBqtJYHQr8AjgBz7DaPFxf21OSpxhXGXGn+oKlsPnN0BSM7j8faNIY3XVgqJwSdaHy1zc+8kq1haUcNPgDK7p1yoomdpuMZEce+LtPTx+D8WuYvyhWpaIM5LP5+Ptt9/mwIEDdO7cmREjRpCUlBTtaQkhalFNAqMRWuv7gPt+fkApNQ4jSBIiMne5sVpU3b7V8Nn1EJ8GE/4BccHVp6sHReGSrDfuL+Wej1dS7PDwzLiuDOkUXEwvzmYm8STaezi9TkrdpWjZRT4rOBwOYmNjMZvN9OjRg4YNG9KunaRWCnE2qMlewEMhHnvwVE9EnIE8DiOvqLoD6+DT3xlBz+XvQ0KT4DHVgqISpydkUDRv40Gu/+BH/Br+flXPoKBIAUkx1pMKisrcZZS4SyQoOgv4/X4WL17Myy+/zObNRnWSPn36SFAkxFkk7IqRUmoEMBJorpR6qcpTSRjbakKE53EaJ9CqO5gPn/zGOHU24X1IPCd4TJWgKFKS9b8W7+SN7zbT6ZxEnp+QTVpiYJK1UpASa8NmObFcoJ/zidz+wCKUW4q2sP7wega3kN5XZ5Jdu3aRm5vL/v37ycjIkDwiIc5SkbbSDgBrACewtsrjpcCfanNSop7zusAZIig6vMUIisx2Y6UoOT14jC3uaAJ2uCRrt9fPM9M3kLtqL8MyG/PwxZ2Djt2bTYrUk0iyDpdPNHf3XJ5b+hxN4pvQv3l/LKGa2op6Z9asWSxcuJDExEQuv/xyMjMzpeCnEGepsO/qWusVwAql1L+11iEKzwgRgtdtFHCsnh1duA0+uRaUyQiKUloGX1slKPL4/BRWuINvU+7mT5+t5qddRVw3oA3XDWwT9AFmMxtJ1ifa3iNUPpFf+3l/7fv8e/2/6dywM68OeVWConpOV35zKaVo1KgRffv2ZfDgwdjtIVrQCCHOGjV5Z2+ulHoS6Awc7c+gte5Qa7MS9ZPPA44jwUFR0S74+Frw+2DiB9CgbfC1VYIit9dPkSM4KNp6sIy7P17JoVI3f7mkC8O7NA26TazNTNJJ5hNVeCsCH/OU8fSSp1m8dzGj2ozi9u63kxaXdsKvIaJv37595Obmkp2dTU5ODt27d4/2lIQQdURNAqP3gSeAF4BRwG+QHCNRnc8DFSGCopI98Mk14HXC5R9AwxBJrFWCIpfXR3GFJyjNeeGWQzz0xRpiLGb+9useZDUPrneUGGMhznZiqzjh8ol2le7i4QUPU1BWwO3db2dsxljZYqnHXC4X3333HUuXLiU2NpaYmBC9+ISoB3788cfGFovlHSCLmh2kEgY/sMbr9V7Xs2fPA6EG1ORTJE5rPUMp9YLWegvwkFLqh1M6TVG/+byVQVG1eLl0n7FS5CozjuSnheg2XiUocnp8lDgCgyKtNR8u28UrczbRrnECL1yeTZOkwA8zBSTHWYOKOdZUuHyiJXuX8OTiJ7GarTw/+Hmy07JP6P6ibti4cSNff/01ZWVl5OTkMHToUGJjY6M9LSFOiMVieadp06aZaWlphSaTSY7M1pDf71cHDx7svG/fvneAsaHG1CQwcinjV+QtSqkbgT1AcKEYcXby+yq3z6oFRWUHjJwixxEY/w9o0iX42ipBkcPto8TpCXja6/PzwsyNfL5iD4M7pPHnsZ2DVoRMSpEaZ8Vygu09QuUTaa3574b/8t6a98hIyeDx/o/TJC5ESQFRr5jNZhITE5k0aRLNmzeP9nSEOFlZEhQdP5PJpNPS0or37duXFW5MTQKjO4EE4HbgSSAZ+O2pmaKo1/x+Y6XIX62Ja8Vh4/RZ2QEY9w6c0y342ipBUajCjWUuL/d/upql249wTb9W3Dg4A1O1LSyr2UTKSSRZh8oncngdvLDsBb7f/T1DWwzl7py7ibH8skKlUMRb46UdSD3g8Xj44YcfUEoxZMgQMjIyaNu2rWyFijOFSYKiE1P5/1vYN/FjBkZa6yWVfy0FrgJQSoU4Zy3OKn6/EQD5qxVddBQaQVHJHrhsCjTvEXxtlaCozOWl3BV4j0NlLu788Ce2HCzn4Yszubhbs6BbxFjMJMVaTuhDLlw+0b7yfTyy4BG2Fm/l+m7XM7HDxID7m5SJZFsyVvOJJ3eL0yM/P5/p06dTVFRE9+7d0VqjlJKgSAhxTBEDI6VUL6A5MF9rfUgp1QWjNchQQIKjs5XWxhZZ9aDIWWxUtC7cDpe+BS16B19bJSgqcXpwuANXm3YeruD2/62gqMLDi5dn0zcjuFVIvN1Cgv3Ekqw9fg8lrhJ8OvB1VxxYweOLHsev/Tw18Cl6Nw2cu8VkIdmWjNl0YnlM4vQoLi5m2rRp5Ofnk5aWxjXXXEPr1q2jPS0hzkhms7ln+/btHT6fT7Vr187x0UcfbU9MTDzpw1m7du2yXH311a0LCgpsXq9Xpaenu+bOnbu5efPmXadOnboxOzvb9fPY3/72ty2aNWvmPu+88yquuOKKjPT0dLfD4TA1atTIc/fdd++74oorio/39cMuJSmlngb+DVwJTFdKPQh8B6wE5Kj+2UprY/vMF5gPhKsUPr0ODm+Csa9Dq37B11YJioodwUHR2oJifv9BHk6Pjzev7BEUFCkgOdZ6wkGR0+ukyFkUEBRprfls02f8cd4fSY1J5Y0L3ggKiuxmO6n2VAmK6gG3282OHTsYNmwYN9xwgwRFQtQiu93u37Bhw7pNmzattVqt+sUXX6xxHROPxxP2ufvuu6/50KFDS/Lz89dt2bJl7XPPPbcH4NJLLz3ywQcfHC1J7/P5yM3NTb366qsLAXJycsrWr1+/bvv27WteffXVnffcc0/LL7/8MvF4/12REiUuAbK11pcDw4GHgYFa6xe11hURrhNnKq2NrTJf4BYU7jL4/Ho4uB4ufgXaDAq+tjIo0lpTVOHG6QkMihZuOcTN/15OnN3MlKtz6NwssIO5SSlS421BFa5rKlS/M7fPzfN5z/PGT29w3jnn8frQ10lPDFwIjbfGk2xPli2YOmzbtm3Mnj0bgLS0NO6880769++P2SyBrBCny4ABA8o2b95sz8/Pt7Vv3/7oaZtHHnmkyV133dUMoHfv3h1vvfXW5r169er4xBNPNCkoKLCMGDEiIysrKzMrKytz5syZ8QD79u2ztmjR4ugHTZ8+fRwAV1999ZHPP//8aGA0bdq0xPT0dFeHDh2qfShBv379HPfee2/B66+/ftyHxSL96u3UWjsAtNZHlFIbtNb5x/sC4gziKDTafVTlqYDPb4S9q+CilyBjaPB1AUFRcN+z3FV7eTJ3Pe0aJ/DXX2XTMCGw8rDFpEg5wfYe4fKJDjkO8ejCR9lwZANXd76aqzpfFZBQrVAk2ZOwm6UKcl1VWlrKzJkzWbNmDampqfTv35/Y2FhsNlu0pybEaXXvJytbbNxXGncq79mhaWLF8xOyd9VkrMfjYcaMGUnDhw8P0TU8UFFRkXnZsmX5AGPGjGlz11137R8xYkTZpk2bbCNGjGi/devWtbfccsuBa6+9tu3f/va3ivPPP7/kpptuOty6dWtPnz59HCaTiUWLFsX27dvX8Z///Cd1woQJR8K9Vu/evSteffXV4ErAxxApMGqrlPqs8u8KaF3la7TW4473xUQ9FjIocsKXt0DBchj1HHQYEXxdZVDk92uKHB48VYIirTUfLNrBm99voVfrVJ4Z3y1om8xuMdp7nMiKjdfvpdhVHJRPtPbwWv688M9UeCp4rN9jDGg+IOB5kzKRbE/GapIk67rI7/ezbNkyvvvuO7xeL4MHD6Z///5YrfLfS4jTyeVymTp16tQZoE+fPqV33HHHoR07dkT8QbziiiuOBjILFixI2rRp09FiYmVlZebCwkLT+PHjSwYMGLD6888/T54+fXpyz549O69evXpts2bNvOPGjTvyr3/9q0FOTs6eWbNmpTz//PMF4V5LVy84XEORAqPx1b5+/YReQdR/zmIjCKrK64avb4Odi2Hk09Dp4uDrqgRFhRVuvFWawfr8mpdnb+SjvN0M79yER8Z0xlqtFlGczUziCbb3CFWfCGDqtqm8uvxVGsU24rlBz9EmuU3glE02kuxJchy/DnM6nXz//fekp6czatQoGjYMTtAX4mxS05WdU+3nHKOqj1ksFu33//ILsNPpDHgzrZqcrbUmLy9vfUJCQlAE06RJE9+NN9545MYbbzwyZMiQdjNnzky49tpri6655pojI0eObD9kyJDSjh07Opo3b+6tfu3Pli1bFteuXbvj7vUa9t1faz0n0p/jfSFRT7lKwV0tpcznhm/ugO0/wIWPQ+dLg6+rDIp8fs2RakGR2+vn4S/W8FHebib3bsljl3QJCooSYywnHBSFyify+r28uvxVXsx7key0bN4c9mZQUBRriSUlJkWCojqooqKCefPmobUmLi6OG264gSuvvFKCIiHqmPT0dO+RI0cs+/btMzscDjVjxozg/k2VBgwYUPLss88ezQFauHBhLMBXX32VWFpaagIoLCw07dixw96mTRs3QJcuXVwpKSm+hx56KH3ixIlht9GWLFkS+/zzzze75ZZbQrb9iETag4vw3BVGO4+q/F6Yeg9s/Q6GPgJdLw++rjIo8vr8FFZ48FdZzixzern3k5Us31nE7Re048o+rQIuPZn2HuHyiYpcRTy+6HFWHlzJxA4Tua7rdQEnzBSKBFsCsRZpD1HXaK1ZsWIFs2fPxul00rZtW9LT00lJSYn21IQQIdjtdn333Xfv7d27d2Z6eror0orNlClTdl133XUtO3To0Nnn86k+ffqU9uvXb+eyZcvi7rzzzpZms1lrrdVVV111aPDgwUd/Q58wYcLhp556Kv3KK68sqnq/vLy8hMzMzM4Oh8PUsGFDz/PPP7/zkksuKT3ef4M60T24aMnJydF5eXnRnsaZz+M08oqq8vtg2r2QPxXOvx96XBN8XWVQ5PH5KaxwB/SUPVjq4g8f/sS2Q+U8cnFnRmYF5sSZlCIlzhq0elQT4fKJNhZu5NGFj1LkLOLunLsZ1mpYtdeUoo111b59+8jNzWX37t20bNmS0aNH06SJtGYRZx+l1I9a65yqj61cuXJ7dnb2oWjNqb5buXJlo+zs7NahnqvxipFSyq61dh17pKj3vG5wFgU+pv0w8wEjKBp4T8SgyO31U1ThDsju2XG4nDv+9xPFDg8vTczmvLaBWyAnc/IsXD7Rtzu/5YW8F0iyJfHK0FfokBpYfkuKNtZdWms+/fRTKioquOSSS8jOzpaSCUKI0+KYgZFSqjfwLkaPtJZKqWzgOq31bbU9OREFPq+xUlR1qUf7YdajsO5L6Hc79Lou+LrKoMjp8VHi8ASEKKv3FHP3RysxKXjzyh5knhNYo+hkTp6F6nfm0z7eXf0uH+Z/SFajLB7t+ygNYhoEjLGb7STZkuTDtg7RWrN+/XratWuHzWZjwoQJJCUlERsrW5xCiNOnJitGrwIXA18AaK1XKqWG1OqsRHT4fUarD12lzpDW8O1fYM3H0OcmOO/m4OuqBEXFjsBqpvM3H+KBz1aTlmjnlUnnkp4aWGoj1mYm6QSSrMPlE5W6S3lyyZMs27eMsRljufncm4OO3SdYE4izntKSH+IkHT58mKlTp7J161ZGjBjBeeedJ9tmQoioqElgZNJa76j2m7Uv3GBRT/n9RqsPf7X/tD+8ACv/Czm/M1aLqqsMiircXkqdgacmv1pZwDNTN9C+SQIvTQwu3JhgtxB/Au09wuUTbS/ZziMLHmF/+X7u6nkXF7W9KOB5KdpY93g8Hn744QcWLlyIxWJh1KhR5OTkHPtCIYSoJTX5VNpVuZ2mlVJm4DZgY+1OS5xWP7f6qN4Udtm7kPcuZE828oqqbztVBkXlLi9lLm+V22neX7idt+ZupU+bBjw9rmtAAKSApFjrCbX3cPlclLhKgvKJFuxZwNNLnybGHMOL579IVqOsgOfNykyyPRmLSQ5i1iVff/01q1evplu3blx44YUkJCREe0pCiLNcTT4lbsLYTmsJ7AdmVz4mzhSh+p+t+RR+eB46joahD4UNikqdHiqqNIP1+TUvzszn0+V7GNmlKQ9dnBlwykwpSIm1YbMc/8mzUPlEfu3nX+v/xT/X/pOOqR15rN9jpMUF9jGUoo11S1FRERaLhYSEBAYOHEiPHj2k2asQos6oSWDk1VpPqvWZiOhwFge3+tg8B2Y9DK36w8hnoHpAURkUFTs8Ac1gXV4fj365lu/yD3Jln5bcOrQdpioBldmkSD2Bk2daa0rcJbh8gfOs8FTw7LJnmb9nPhe2upA7e94ZtE0WZ4kjwSarEHWBz+dj4cKFzJs3jy5dunDppZeSlpZGWlqNG3ILIeqYnTt3Wm6++eaWK1eujLPZbDo9Pd312muv7erWrVvQKfb8/HxbdnZ2VuvWrY/WNvrpp5/WT5kypcGjjz6a3qRJEw9AZmZmxeeff779NP4zAtQkMFqmlMoHPgQ+01ofd7EkUUeFqmq9aynk3glNusKYV8FcrSGnLQ5tT6KkwoPT+0tQVOr0cO/Hq1ixq4g7LmjP5D4tAy8zGyfPTMcZFIVLsi4oK+DhBQ+zs3QnN2ffzLj24wJOmEnRxrpl27ZtTJ06lUOHDpGZmcmQIXJ+Q4j6zu/3M3bs2HaTJ08+/M0332wFo3p1QUGBNVRgBNCiRQtX9TYiAGPGjCn84IMPdtb2nGvimHsLWusM4AmgJ7BaKfWFUkpWkOq7UFWtD6yDL2+G5BZw2Vtgiw98vjIoKqoWFB0odXLD//3I6j3FPH5Jl6CgKMZqJiXu+IMin99HkasoKCjK25/HzbNv5rDzMM8OfJbxHcYHBEUmZSLFniJBUR3x448/8sEHH+Dz+Zg8eTITJ04kOTlslwAhRD3xzTffJFosFv3HP/7x4M+P9evXzzF8+PCyG264Ib19+/ZdOnTo0Pntt99OPZH7r1271j5w4MD2Xbp0yezZs2fHFStWxAAUFBRYRowYkZGVlZWZlZWVOXPmzPhj3et41CgTVWu9EFiolPoz8DLwb+B/x7pOKTUSeAUwA+9orZ+p9vyNwC0Yp9zKgOu11kGRpDjFPE5jC62qwh3w2e/Bngjj34HYat/Htjj8NiMo8vh+Oc6/7VA5d/xvBaVOL3/91bn0bhNYLyjebiHhBE+eFbmK8FcpHaC15uONH/P2qrdpldyKx/s9TrOEZgHXSdHGusHv91NRUUFCQgKdOnWirKyMfv36YbVKhXEhTrkvbmnBgXWntgZJ484VXPpGxOa0q1atis3Ozq6o/vgHH3yQsnr16tj169ev3bt3r6V3796Zw4cPLwPYtWuXvVOnTp0BevXqVfZ///d/OwG+/vrr1E6dOiUA3HTTTfvvuOOOw9ddd12rKVOm7Ojatavr22+/jb/ppptaLl68eOMNN9zQ4q677to/YsSIsk2bNtlGjBjRfuvWrWtP1T+9JgUeE4BLgElAJvAl0K8G15mBN4ALgd0YW3JfVQt8/qO1fqty/FjgJWDk8f4jxHEIVdW67AB89jvjqP7EdyHxnMDnK4OiwmrNYFftLuLuj1ZiMZt469c96dg08ehzJ3PyzOPzUOwuDgiKXD4XL+a9yJydcxjYfCD39b4vaEVIijbWDbt37yY3NxeLxcJvf/tb4uPjGTx4cLSnJYQ4TX744YfEiRMnHrFYLLRo0cLbp0+fsvnz58fl5OQ4arqVVlxcbFqxYkXC5ZdfnvHzY263WwEsWLAgadOmTUc/AMrKysyFhYWm1NRUP6dATX6VXwN8DTyntf7hOO7dG9istd4KoJT6H0aAdfT/EK11SZXx8UD9atxW34Sqau0shs+uM2oYXf5PaNA28BprLL7KoMhXJSiat/EgD32xhrREO69O6k7z1F+ClJM5eRbqOP5hx2EeXvAwGws38tus3zK50+Sg4EeKNkafw+Fg9uzZLF++nMTEREaMGBHtKQlxdjjGyk5t6dq1q+OLL74I2iY7FT1YfT4fiYmJ3lBBlNaavLy89QkJCbUSM9Tkk6ut1vq24wyKAJoDVf9j7a58LIBS6hal1BbgOSBEBUFQSl2vlMpTSuUdPHgw1BBxLKGqWnsc8MVNULgNxr4OTbsGXmOx47UlcaQ8MCj68qc93PfpKjLSEnj76pyAoMhsUjSIO7GgyOF1UOwqDgiKthRt4dY5t7KjZAeP93+cKzOvDJlPJEFRdO3du5fXX3+dFStWcN5553HLLbfQpUsXWb0T4gw2ZsyYUrfbrV588cVGPz82d+7cuNTUVO8nn3zSwOv1UlBQYFm6dGnCwIEDy4/n3g0aNPCnp6e733vvvVQwtucXLVoUCzBgwICSZ599tvHPYxcuXHhKE0rDrhgppV7UWt8NfKqUCorKtNbjjnHvUO+Ioe7zBvCGUmoy8BAQ1J1Uaz0FmAKQk5Mjq0rHK1RVa58HvvkDFKyAi/8KrartjlrseG3JHKlwH11g0lrz3oLtTJm3lb5tG/LUuCzibL98C1nNJlJO4OQZQLmnnHJP4M/Nkr1L+MvivxBniePlIS/TPrV9wPNStDH6vF4vFouFtLQ0MjIy6N+/v7TyEOIsYTKZ+Oqrr7bcfPPNLV5++eWmdrv96HH9srIyc2ZmZhellH7sscd2t2zZ0pufn2879l1/8d///nfr73//+1bPPvvsOV6vV1122WVH+vbt65gyZcqu6667rmWHDh06+3w+1adPn9J+/fqdshNtKtySl1Kqt9Z6qVLqglDPa63nRLyxUn2BP2utR1R+fX/ldU+HGW8CCrXWEY+r5OTk6Ly8vEhDRFVaG0FR1QKO2g/T74f1X8KwP0O3aocMzTa89hQKHV78ld8fPr/mhRn5fLZiD6O7NuXB0ZlYqhRujLGYSYq1nNAKQYm7BKfXGfDYF5u/4I0Vb9A2pS1PDHiCtFgp2liXuFwuvv/+ezZu3MiNN94oSdVC1CKl1I9a64BeOStXrtyenZ19KFpzqu9WrlzZKDs7u3Wo58L+qq21Xlr510yt9etVn1NK3QpEDIyAZUB7pVQbYA9G8vbkavdpr7XeVPnlRcAmxKlVvaq11jD3WSMo6ndHiKDIii8mhcKKX4Iip8fHI1+uZe7Gg1zdtxU3n58READF2cwknkAj2FCFG33ax99++hufb/6cfs368UCfB4KSrKVoY/RorVm3bh0zZsygtLSUnj174vefknxHIYSoE2qyB/Fb4PVqj/0uxGMBtNbeygBqBsZx/fe01muVUo8DeVrrr4BblVLDAA9QSIhtNHESQlW1Xvp3WP5P6H4V9Lkx8DmTBV9MasBKUYnDwz0fr2TV7mLuHNaeSb0DaxQlxViJtR3/yTO/9lPsKsbj9xx9rMJTwRNLnmDJ3iVM6DCB67tdj1n9cm+FItGWSIwl5rhfT5w8p9PJxx9/zNatW2natCkTJ04kPT092tMSQohTKlKO0a8wVnnaKKU+q/JUIlAU+qpAWuupwNRqjz1S5e93HNdsRc2Fqmq96kNY8DJ0GgPn3x/Y/8xkxl8ZFP2caL2/xMkf/vcTuwor+MulWVzY+ZfcEQUkx1mxW44/KPq5cKNPVykSWXGAB+c/yPaS7dzR4w7GZowNuMakTCTbk7GaZMvmdNNao5TCbrdjsVgYNWoUOTk5mEyyjSmEOPNEWjFaChwG0jHqEf2sFFhRm0y4kb8AACAASURBVJMSJylUVeuNM2DOY9B6EIx4KrD/mTIZQZHTdzQo2lvs4OZ/L6eowsPLvzqXnNa/FG40KUVqnDUgx6imPH4Pxa7AGkX5R/J5aMFDuLwunhrwFL2a9gq4xmqykmxPlnyiKNi4cSPfffcdkydPJjExkUmTJslJMyHEGS1SjtE2YBsw+/RNR5y0UFWtdy6GafdA024w5mUwV1l1USZ0bCqFTv/R4o17ix3c9K/llDq9vDG5B52bJR0dbjEpUk6gESyA2+cOOo4/f898nlryFCn2FJ4b+hxtktsEXBNjiSHRmigfxqdZUVERM2bMYMOGDTRq1Ijy8nISE+W/gxDizBdpK22u1nqwUqqQwGP2CtBa6wZhLhXREqqq9b7VRv+zlNZw6VtQtd6PUujYFApdHA2KCoqMlaJSp5fXrugeEBTZLUYj2BP5cHR6nZS6S48GRVprPtr4EW+veptODTrxeP/HaRDzy7eUQhFvjZf6RKeZ1poFCxYwb948AC644AL69u2L2SwtVoQQZ4dIexM/t79uBKRV+fPz16IuCVXV+shW+Px6iE2Bce8Y//szpdAxKRS51NHeZ1WDotcnBwZFsTYzKXG2EwqKKjwVlLh/qWbt9Xv5649/ZcqqKQxKH8SL578YEBRJ0cboUUpx8OBBMjIyuPnmmxkwYIAERUKIkJRSPS+99NKjy/wej4fU1NTsIUOGtAN49dVXG6ampmZ36tSpc0ZGRpeqhSDrskhbaT8ngbQACrTWbqXUAKAb8C+gJNy14jQLVdW6dB98eh2gYPx7kBhYdE/HJFPsMeGuFhSVuYygKPOcX4KixBhLQCHH41HmLqPCWxHw9WOLHmP5geVM7jSZ32T9JiB3SJrAnn5lZWXMmjWLvn370rRpU8aOHSvBkBDimGJjY/35+fmxZWVlKiEhQX/++edJTZo08VQd83MPtD179liysrK6TJw4sahFixbeaM25JmqSzfoFoJVSGcAHGI1k/1OrsxI15/cbK0VVq1o7Co3+Z65iGPc2pLYOvCYmmWKPGZc3OCh67YpfgiIFJMdaTygo0lpT7CoOCIoKygq47dvbWHVwFff2upffdf1dQFBkN9tJtadKUHSa+P1+li5dyuuvv87atWvZu/f/27vv+Kiq/P/jrzOTmUmdVEpCS2gJvfcmCAiCoCKCi13Xr2XtKDZAUSxYsMAq6PKzu64srihFqoB0kN6k1wAhkJ7p5/fHDBAwwCSk5/N8PHiYmblz72fOIzJv7j33c5IBJBQJIfx27bXXpv/www8RAN99913UkCFDTue3XY0aNVy1a9e279mzp0Ddr0uDP994Hq21Uyl1M/C+1vpDpZTclVYWaO2dU+TOE9CdOfC/ByHtINz0KVRrcuF7Aq2ku0zYXd4gddlQVMjb8c+GIofnfGPJbae2MXr5aDzaw1vd36Jl1ZYXvEcWgS1ZR48eZdasWSQnJ1O3bl2uv/56oqOjS7ssIUQhjF4+utaeM3uK9C/Q+pH1c17t8uoVF6e94447To8dOzZ22LBhaTt27Ai+7777UlesWPGXDrzbt283Hz582NK4cWN7fvspS/wJRi6l1FDgDuBG33PSTKYssGdc2MDR7YCZj3knXA/8AGp3vHB7SyjpbjO2PKHooa//IMdxUShSEBFUuIVgPdpDmj0Nl+f8mdJFhxYxYe0EqgZXZXzX8dQKq3XuNYMyYDVbMRvL/D8iKpQ///yTrKwsbrnlFho3bix3mwkhCqVDhw65R44csXz66adRvXv3Tr/49Z9//jkyKSkp1Gw2e95///2D1apVc+e3n7LE387XDwMTtNb7fEt8fFe8ZYkrctkvbOB4dv2zg79Dn1ehQZ8LtzeHkKEDsTnzCUV/a0VS9fOhKDLYjKkQPYpcHhdp9rRzPYq01ny942s+3/Y5zWKa8UrnVwi3nF8KTxaBLTlaazZu3EhYWBj169ena9eudO7cGYvFUtqlCSGukj9ndopTv3790saOHVtr3rx5u06ePHnBX+hn5xiVVm2FccVvJK31VqXUY0B9pVQSsEdrPb74SxOX5PFc2KtIa1g8HnbNgq5PQbOhF25vCiKTIHId3lB09Iz38tnFoeiqGje6naQ7zjdudLgdvLvuXRYcWkCfOn14qs1TF5wVshgtWM1WOVNRAk6cOMGsWbM4fPgwzZo1o379+rLoqxCiyDz00EOnwsPD3e3bt8/95Zdfwkq7nqt1xWCklOoGfIV3IVgFVFdK3aG1Xl7cxYlLsKdfONl61T9h4zfQ5m5o9/cLtzUFkmUIJcfuvbR19EwuD32znlyHu8hCkd1tJ8N+/nb8dHs6Y1aMYeuprdzT5B5GNBpxQQAKMYUQYgop8HFEwdjtdn777TdWr15NUFAQgwYNomXLlld+oxBCFEC9evWco0ePPlnadRQVf65hTASu11pvB1BKNcIblNoWZ2HiEpy53u7WZ236DlZ+BI1vhO7PXrj+WYCFbBVK9sWhyOlm0t9ak1jdG+yNBkVkIbtZ21w2MhznOzcczjzMC8teICU3hRc7vEiv2r3OvaZQWC1WLEa5fFMSdu7cyapVq2jdujW9e/cmKCiotEsSQlQgOTk5f7kRa+DAgZkDBw7MBHjsscdS8S4tVq74E4zMZ0MRgNZ6h1JKZsqWBo8bbHnaR+2aAwvHQd2e0Pe1C9c/M5rJMYaRZfeeWTpyJoeHv/nDG4puOx+KAnyhyFCIUGR32y8IRRtObuDlFS8TYAjg3R7v0iTm/B1xMp+oZKSmpnLq1CkSExNp3rw51apVo3r16qVdlhBClBv+fEv9oZSagvcsEcAIZBHZ0mFLP9/E8fgWmPss1GgNAyZC3sBhNJEbYCUzTyh66Os/sLncTP5baxpWK5pQlG4/P89p7v65vLf+PWqE1mB81/HEhcade81sMGO1WGUR2GLkdDr5/fffWb58OaGhodSvXx+j0SihSAghCsifYPQg8BjwLN45RkuBj4qzKJEPR/b5W/Nz0+CXxyE4BgZNAlPg+e0MAdhM4WTYLh+KTEYDkcGFW/fs7Jwi8N6eP23rNL7b+R2tq7ZmbKexhJrPt7AIDgi+4LEoert372bOnDmcOXOGZs2a0adPH2nSKIQQhXTZYKSUagbUA37UWk8omZLEX7hdYM/0/qw9MHcUZKXAsG8gKPL8dgYjDnPEuVB0+LT38tnFochsNBBxlaFIo3F5XLy15i0WHV7EgIQBPNb6sXOXyhSKMHMYgQGBV9ijuBonTpzg22+/JTo6mjvvvJOEhIQrv0kIIcQlXTIYKaVeAO4D/gDaKaXGaa2nlVhl4jxb+vnFYVdPgf1LoNcYiG1+fhtlwB0YSVquG403FD30zR84XJ4LQpElwEB40NWHIofbwbiV41iZvJL7mt7HbUm3ndunQRkIt4RjMsgt4cXB7XZz6NAhEhISqFatGsOGDaNBgwZylkgIIYrA5c4YjQCaa62zlVJVgNmABKOSZs/0drQGOLgCVnwISQOhxW3nt1EGdFAkaTYPWl8Yiib9rdW5UBQYYMQaFHDVoSjXlcvYFWNZf2I9j7Z6lBvr33huO5PBRLglXOYTFZMDBw4we/ZsTp06xaOPPkpkZCRJSUmlXZYQQlQYl/v2smutswG01ilX2FYUB7cT7FnenzOPw+ynIboe9Bl3/rZ8pSAoknQHuDyaExk2Hj57pmhEnlBkMhJeBJfPspxZPLfsOTac2MAz7Z65IBQFBQQRYYmQUFQMsrKy+PHHH/niiy9wOp0MHz6cyMjIK79RCCGKidFobJOUlNS4QYMGTXr16lX/1KlTRoBdu3aZlVJtxo8fX/XstnfeeWftDz/8MBpgyJAh8TVq1GiWmJjYOD4+vulNN90Uv3///jJzieFy32B1lVIzfH9+BOrleTyjpAqstLT2TrIG7xmjX57wTr6+4UM4u9iqLxRluBR2l4dMm5Mnv994bkHYBlW9oSjIbCQ8qHC/cw6341woSren8+ySZ9mRuoMXOr5Av/h+3jJ884nCzGHSyboYOJ1OPvnkE7Zu3Uq3bt14+OGHadiwYWmXJYSo5CwWi2fnzp3bd+/evS0iIsL19ttvVzn7WlRUlGvKlClVbTZbvl8Kr7322pFdu3Zt37dv39aWLVvm9OzZM/FS25a0y11KG3LR40nFWYi4iD0Dzi7EuvRtSN4IA9+HqLrntwkMJ8djJNfhwuHy8Oz0zRxIzeH9YS3P9SkKNhsJCyx8KEq3p6PRnLad5tklz3Ik6wgvd36ZznGdAd98InM4JmOZCfsVRmpqKtHR0ZhMJvr06UONGjWIiYkp7bKEEOIvOnbsmL158+ZzXWSjoqJc7dq1y5o8eXL0008/fepS7zMYDIwdO/bkzz//HDl9+vTw22+/Pa1kKr60SwYjrfXCkixE5JF3gdhds2HDV9DqTmjY7/w2ljDsykxmrhOP1oz7ZTt/HErj5UGNaZ8QBUCIJYBQS+EaKuYNRSdzTjJyyUhSc1MZ33U8baq1AbyhKMISIU0bi1hubi4LFy5k/fr1jBgxgvr169OiRYvSLksIUUYde+HFWvbdu4OLcp+WBg1y4l4f79fitC6Xi8WLF4fdd999FwSgMWPGJPfv37/B448/fslgdFbz5s1zduzYUSZuY5ZvtLIm7wKxqXth3ksQ1wq6jzy/jSkIV0Aw6dneSdmTFu1h/vYTPHxNPfo3jQUg1BJASBGEomNZxxi5ZCRZjize7P4mzWKaARBgCJD5REVMa82mTZuYP38+ubm5dOjQgVq1apV2WUIIkS+73W5ISkpqfPToUXPTpk1zbrzxxoy8ryclJTlatmyZPWXKlKgr7UufvfO6DJBgVNacXSDWkQ0/PwYBgTDgfTi7Mn2ABY8lnDPZDjTw/drDfLP6EENa1+DOTnWAogtFBzIO8OySZ3F4HLxzzTs0jPTOazEbzIRbwmU+URH7z3/+w86dO6lZsyYDBgyQrtVCCL/4e2anqJ2dY5Sammrs27dv/TfffLPqSy+9dMFismPGjDl+66231uvQoUPm5fa1ZcuW4N69ex8v3or94/c/95VSsvJncTu7QKzWMH8MnNkPA96FsGre140mdGAEab7LZ4t2nmTi/D/p0bAKT/dNRClFsNlYJKFo95ndPLX4KTzaw8RrJp4LRRajRUJREXI4HHg83mVeGjduzKBBg7j33nslFAkhyo3o6Gj3hx9+eGjy5MnV7Hb7BV8OrVq1sjVo0CB34cKF4fm91+Px8Nprr1VNSUkxDRkyJCO/bUraFYORUqq9UmoLsNv3uIVSSpYEKWp5F4jd9C3smgWdH4PanbzPGYzeO9BsbpxuDxsOnWHsT9toWiOccYObYDQogopoovW21G08veRpLEYL7/d8n4RwbzfloIAgCUVFRGvN9u3bmTRpEuvXrwegWbNmtGrVSsZXCFHudOnSJbdRo0a5n3322V/6iIwePTr5xIkTFyw+/9JLL9VMTExsnJCQ0HTdunUhixYt2hUYGFgmrqf5c2rhQ2Ag8D8ArfUmpVTPYq2qMjq7QGzyJvjtTUjoAe0f8L7muy0/y6mxudzsS8ni2embqR4eyLtDWxBoMnqbNxZBKNp4ciMv/v4iUYFRvNPjHaqFeM9WhZpCCTYV6dy+Suv06dPMnj2bvXv3Ur16dWJjY0u7JCGEKLCcnJwLFpRftGjRnrM/7969e9vZnzt16pTr8XjWn3383//+90CJFFhI/gQjg9b64EX/inUXUz2Vkz3Teyda7hlvv6LQqtB/Apyd2BwUSa7bQLbdSUqmnSe/30SA0cAHw1sSHmzCbDRgDSrc5TOn23kuFK1OXs3LK14mNjSWt7u/TXRQNABh5jCCAoKusCfhj3Xr1jF37lyMRiP9+vWjXbt2GAwygV0IIcoKf75NDyul2gNaKWUEHgX+LN6yKhGX3dvd2uOG2SMh5xQM/w4CfZdjA8NxYCLT5iDL7uKJ7zeSnuvk49tbExcRhOkqFoR1up2k2dPQaJYeWcr4VeOJD49nQvcJ3ktmKKwWKxajTC+7Wh6PB4PBQFRUFI0aNaJv376EhYWVdllCCCEu4k8wegjv5bTawAlgge85cbU87vPdrVf9Ew4uh97joFpT73PmEFzGQNJyHDjcHp7772b2n8rm3aEtaBRrJcCgiCjkgrB5Q9H8g/OZsGYCSdFJvNH1DULNodK4sYikp6fz66+/EhERQd++falbty5169a98huFEOLyPB6PRxkMhjIxL6c88Xg8CvBc6vUrBiOt9UlgeFEWJfAt+XHGO69o/zJvMGp8IzQb6n3dFIjHHEZajgOPR/ParB2sPXCG0QMb0aleNEaDIjLYjMFwdaHo570/88EfH9CiSgte6/oaQQFB0rixCLjdblatWsWSJUvQWnPNNdeUdklCiIpla0pKSuMqVaqkSzjyn8fjUSkpKeHA1kttc8VvPqXUp8BfBl1r/cDVlVfJ2dK9i8SmH/FeQotpANeO9U60NprRlnDSc524PZp//raXuVuP83/d6zKweRwGVTSh6Ic/f+CTTZ/QIbYDYzuNxWK0EGAIINwcjtFgLIYPXTkkJyfz448/kpKSQmJiIv369SMiIqK0yxJCVCAul+v+48ePf3b8+PGmyCLvBeEBtrpcrvsvtYE/pwQW5Pk5ELgJKJVmUhWGI+d8z6KZjwIaBk0CUxAYAry35dvdONwepq8/wpcrD3Jjyzju6RKPUhAZbMJ4FaHIoz18teMrvtj2Bd1rdueFDi9gMpgwG8xYLVbpZn2VTCYTHo+H4cOHk5iYWNrlCCEqoDZt2pwEBpV2HRWRP5fSvs/7WCn1FTC/2Cqq6NxO7wKxWsPCsZCyA278BCJqe+9CC4ok2+nB5nSzZFcK7/y6i24NYnimXyIGpYgIMhNgLHhwyRuKpm6Zyn92/Ye+dfoysu1IjAYjFqMFq9kqPXQKwePxsH79epKTkxk0aBAxMTE88sgjMpZCCFEOFWYSSQJQp6gLqRQ8Ht+8Ig2bvoPtP0HHR6DuNed6Fdk8iiy7k81H0hj901Yax1l5dXBTTAaD99b8gMKFonRHOm7t5qMNHzFz70wG1RvEo60exaAMBAUEEWaWO6QK4+jRo8yePZtjx46RkJCA0+nEZCrchHghhBClz585Rmc4P8fIAJwGnivOoiosW5r3TrRjG+C3N7xNHDs94n0tMBynCiAjx8HB1Gye/mETVcIsvDu0BUFmI9YgE5aAgs/7cXq8ocjpdvLOuneYd3Aetza8lQeaP4BSihBTCCGmkCL+oBWfzWZj4cKFrFu3jtDQUIYMGUKTJk0kEAkhRDl32WCkvH/LtwCO+p7y6LK0BG55craJY3YK/Pw4hFU/38TREorbGEhatoNTWXYe//dGjErxwfCWRIaYsQaaCDQVMhTZ07G77byx+g2WHFnCXU3u4o5Gd6CUksaNV8Hj8bB9+3Y6dOjANddcQ2BgYGmXJIQQoghcNhhprbVS6ketdZuSKqhCOtvE0e2EX570zjG6+XtvE8cAC9ocSlq2g0ybkyf/s4kzOQ4+HtGGmpHBhFoCCDIXPhTZXDZeXvkyq5NX83/N/49bE2+Vxo2FdPLkSdauXUv//v0JDg7msccew2KRMRRCiIrEnwkra5RSrQuzc6VUP6XULqXUHqXUXy6/KaWeUkptV0ptVkotVEpVvLlLeZs4LnsHjq6DPq9ClUTvHWiBEWTkurA53bzw4xb2nMhi/E3NaBxnJdhsJMRS8GlgZ0NRtjObF35/gdXJq3mi9RPcmnjruR5FEor853A4mDdvHp988gnbtm3j9OnTABKKhBCiArrkt65SKkBr7QK6An9XSu0FsgGF92TSZcOSb/mQyUAf4AiwVik1U2u9Pc9mG4C2WuscpdRDwARg2FV9orIkbxPHnb/AH19Aqzug0Q3nJlvnuDzkOl28Pmcnq/ad5oXrk+haP4Ygs5GwQiwKezYUZdgzeOH3F9iRuoNR7UbRN76vNG4sIK01O3bs4NdffyUjI4NWrVrRu3dvgoNlMV0hhKioLvcNuQZoDdxYyH23B/ZorfcBKKX+DQwGzgUjrfXiPNuvAm4v5LHKJnuG9/JZyi6YNxpqtIXuz3pfC4zAiYEsm4OpS/cxa3My93VNYHDLGgQGGLFeRSg6YzvDqKWj2J++n9GdRtO9Zndp3FgIHo+HRYsWERQUxC233EKtWrVKuyQhhBDF7HLBSAForfcWct81uLAR5BGgw2W2vw+Yk28hSj0APABQu3btQpZTwhw53j+2DG8TR0soDJwIRhOYQ/AYLaRlO1i08yTTlh9gYPNY/t4tAbPRgDWo8JfPUnJSeGbpMyRnJTOuyzg6xHbAZDARbgmXxo1+cLlcrF69mnbt2mE2m7n99tuxWq0YDDJ2QghRGVzuG7iKUuqpS72otX7vCvvO777lfO9oU0rdDrQFelziWFOBqQBt27Yt+3fFnWvi6IG5z0JmMtz6BYRUgQALBFrJyHFw4FQ2437ZTuNYK6P6JWEOMBIRXPAeOGdDUXJWMiOXjOS07TSvd3udVlVbSePGAtizZw+zZ8/mzJkzhIWF0bx5c1nKQwghKpnLBSMjEEr+AccfR4C81x5qAscu3kgp1Rt4EeihtbYX8lhlh8fjnWytNaz6GPb9Br1GQ1xrMBghMIIsu4u0HCfPzdiMyWjgjZubEWw2EhFU8FDk8rhIt6dzKOMQzyx5hmxXNhN6TKBJdBNp3OinjIwMfv31V7Zv3050dDR33HEHdevWLe2yhBBClILLBaNkrfW4q9j3WqCBUioBbx+k4cDf8m6glGoFTAH6aa1PXsWxyg5bGnhcsG8JrJwEjQZDi7+dm2xt92iybE7emLODfSnZvD+8JXERQYVaFNajPaTZ09ibtpdnlz6L2+Pm3R7v0iCygTRuLIBffvmF/fv306tXLzp16kRAgExOF0KIyuqKc4wKS2vtUkr9A/gV79mnaVrrbUqpccA6rfVM4G28Z6V+8J0pOaS1Lr+L4tmzvD2L0g7BnGegShL0ftkbigLDcasA0nPsTF9/hF+3neD/utelU91oIoJNhQ5Fu8/sZuSSkZgMJt7r+R7x1nhp3OiHgwcPEhkZidVqpV+/fiiliIyMLO2yhBBClLLLBaNrr3bnWuvZwOyLnhuT5+feV3uMMsNl93a3duZ6J1ujYNBHYAoCcwiYgkjPdrD5SDrvL9hNl/rR3N0lHmuQCVMBF4XVWp+7fPbcsucwG828d8171AytKY0bryA7O5v58+ezadMm2rZty4ABA4iKiirtsoQQQpQRlwxGWuvTJVlIuXa2iaPWMH8MnPoTbpoK4TXPTbbOtDk5mWHj+RlbqGq18PINTQixBBRqqY8MRwYnck7w7NJncXqcvN/zfWqF1SLcHI7JWPDb/CsDj8fD+vXrWbRoEQ6Hg65du9KtW7fSLksIIUQZI5MpioIt3XsH2savYefP0PlxSOh2brK1zekm0+Zi9E9bSc9x8tldbYkOtRBWiK7WGY4MUnNTGbV0FGdsZ3inxzvUDa8rjRuvYOnSpSxZsoSEhASuv/56YmJiSrskIYQQZZB8k14tl8N7Ge3oeljyFtTtCR3+79xkazeKDJuDKUv3svbAGV4c0Iik2DDCC3EHWpYjizRbGi8uf5FDGYcY33U8TWKaEGmJlMaN+cjNzcVmsxEZGUm7du2Ijo6madOm0rpACCHEJUkwulr2DMg6Cb88AdYa0H8CKAMEhqMNAaRlO1iyK4UvVhxkcMs4BreIIyLIjLGAk61znDlkODJ4ddWrbDu1jZc6vkT72PZEWCIkFF1Ea83mzZuZP38+UVFR3HPPPYSEhNCsWbPSLk0IIUQZJ8HoajhywGnzhiJ7Fgz5F1jCwBwMpiAycp3sP5XNKz9vJ7F6GE/3bUhoYADmgIJNtra5bGQ4Mnh77dusSl7FE62foGetnoSbw+Xy2UVOnjzJ7NmzOXjwIDVq1KB///5yhkgIIYTf5Fu1sLT23oW2+d9w7A/o9xbENPRNtg4n1+EmLcfBczO2YFDw5s3NCA80E2wu2JDb3XbS7el8vOljFhxawD1N7mFQvUGEW2Si9cX27dvHN998g8ViYeDAgbRu3VpCkRBCiAKRYFRY9kzIPgXLP4BaHaHRoHOTrZ1uDxm5Dt6au5O9J7N4b1gLakcFF3gNNKfbSYY9g293fsuM3TO4ucHNjGg0AqvFitloLqYPVr5orcnOziY0NJTatWvTqVMnOnfuTHBwcGmXJoQQohySYFQYHjc4c+D397z/7TUaDAYIikQrRXqOgxkbjjJ7y3Hu75pAl/oxBZ5s7fK4SHekM3PvTKZtnUbv2r15qMVD0qcoj9OnTzNnzhxSUlJ4+OGHMZvN9O5dcVpjCSGEKHkSjArDngHHNsLW6dDmXoiuBxYrGE1k5DjZciSd9+b/Sce6UdzbNQFroImAAjRxdHvcpNnTWHxoMR/88QEdYjvwTLtnsJqt0tEacLlcLF++nGXLlmE0GunZs6cs4yGEEKJIyLdJQbkcYM+GReMgpCp0etjX3TqYHIeL4xm5PD9jCzGhFsYNaoo1yFSgJo5nl/pYc3wNr69+nSbRTRjTcQxWs5Vgk1weyszM5PPPP+f06dM0bdqUvn37EhYmC+UKIYQoGhKMCsqeAVt/gBPb4Pp3vGeKLFYcLg/pOU7G/LSN1Gw7n97ZlqpWC6EFaOKotSbNnsa21G2MXT6W2tbajO86nojACELNocX4oco+l8tFQEDAublEAwYMoG7duqVdlhBCiAqmYPeNV3aOHG/Pot8nQs12kDgAAq14UKTnOvl02T5W7z/N030TaVojHGug/3eNnV3/bG/aXp5b9hyRgZG82e1NooOisZqtxfihyja3282KFSv44IMPyMjIQCnF4MGDJRQJIYQoFnLGyF9nb8//faK3Z1Gv0WAKPLc47NLdKUxbfoABzWO5qWUc4UEmDAVo4pjh/4cAjAAAHjVJREFUyOBw1mFGLR1FgApgQvcJxIbEVupQdPDgQWbPns3Jkydp0KBBaZcjhBCiEpBg5C97JiRvgi0/QOs7oUoiBIaTZXdxIDWbV2Zuo0HVUJ69LpHwYDOmAky2znRkciLnBKOWjiLHlcPEayZSx1qHcEt4pezD4/F4mDlzJps2bSI8PJxhw4aRmJhYKcdCCCFEyZJg5A+PGxxZsOhVCI6GTo+CORS7B1Kz7Dz33y14NLw5pBmRIeYCTbbOceaQmpvK88ue50T2CSb0mEBiVGKlDEVaa5RSGAwGDAYDXbt2pVu3bpjN0rNJCCFEyZBg5A97BmyZDsc3eztcB0XgDggmPcfBu/P+ZNeJTN4d2oKEmNACzStyuB2ctp1mzIox7Enbw6tdXqVFlRZEWCIwqMo1/evYsWPMmTOHAQMGUL16dW644YZKFwyFEEKUPglGV+JyQMZxWPYu1GgDjQahA62k21z8tOEYMzcd457O8XRvWIWIIP9D0dleRa+vfp0NJzfwXPvn6BzXmUhLZKUKRTabjUWLFrFu3TqCg4PJysoCkFAkhBCiVEgwuhJ7Bqz40PvfXqPBEkKmy8CWo2d4+9ddtI+P4oHudQs02VprTbojnc+3fc6yo8t4qMVDXBd/HeGWcIwG/y/DlXdbt25l7ty55OTk0K5dO3r27ElgYGBplyWEEKISk2B0OY4c74Trzf+GFn+Dqo2xGYI5fsbGc//dQmSIiXGDmxAebMIcUIDJ1s5Mlh1Zxlfbv+K6+Ou4pcEthJvDMRkq16Kwp06dIiIighEjRhAbG1va5QghhBASjC5Ja+9ZooXjICgSOj+GyxRKWq6Ll2duIyXTzpQ72hAbEUSw2f9hzHXlsjdtL2+seYMGEQ14ovUTRARGYDJW/FDkcDhYsmQJderUoWHDhnTr1o0ePXrIZTMhhBBlhgSjS7Fnwtb/QvJGuO51dEgM6a4Api3fx4q9qTxzXSIta0VgDfR/CJ0eJyk5KYxdPhaDMvBy55eJDorGbKzYd11prdm5cydz584lIyMDo9FIw4YNMRorz2VDIYQQ5YMEo/x43JB5HJa+A7EtoclNZBDC73tO8enSffRrWp1b2tQgItjs99kOj/aQZkvj3XXvciDjAG92e5OE8IQKvyjsmTNnmDNnDrt376Zq1aoMGTKE2rVrl3ZZQgghRL4kGOXHngHLPwBbGlz7GbmGEPafzmX0T1upVyWU5/snERFsxliQztb2DKbvns6iw4u4t+m9dIrrRKip4q9/dvDgQQ4ePEjfvn3p0KEDBkPlueNOCCFE+SPB6GIuBxzbCJu+hebDcVdrTqrDxPMz1uH2aN4Y0ozoUAuWAP8vA2U5slh3Yh1TNk2hc1xnRjQagdVsrbBza/bu3Ut2djbNmzenRYsW1K9fn9DQih8ChRBClH8SjC5mS/d2uA4Mh86PkUUw787/kx3JmUwY0pwGVUMJtfg/bHa3nUOZhxi3chzVQ6ozqv0oIi2RFfK2/IyMDObNm8e2bduIjY2lWbNmKKUkFAkhhCg3JBjl5cjxTrg+uh76vIojpDozNp3kxw1HubNTHXo1qlqgztYuj4vU3FTGrRxHriuXt3u8TfXg6hXuDjS3282aNWv47bff8Hg8XHPNNXTp0qXCnhETQghRcUkwOktryEyGpW9D9ebQ/FYOZhl5b/6ftKwVwYM96hJR0CaO9nQ+3vQx21K3MbrjaJKikgg2BRfzByl5x44dY968eTRo0ID+/fsTGRlZ2iUJIYQQhSLB6CxHlnfCdU4q3PgxuYZQ3l/4J7kON8/1TyIqxEKA0f+JwxmODOYemMv/9vyPWxrewrW1r8VqthbjByhZ2dnZ7N+/n6ZNm1KrVi3uv/9+4uLi5CyREEKIck2CEXhvzz+2ATZ8Dc2Gomu2Zcn+TGZtSeauznVoEmcl0OT/nKBcVy7bU7czcf1EWlRpwYPNHyTcEl4hQoPWmj/++IMFCxbgcrlISEggJCSEGjVqlHZpQgghxFWTYATeCdcLx4ElFLo9xRl3IG/N2UxseCB/71aXsALMK3J6nBzNPMrYFWMJM4fxUseXiAiMIMBQ/oc6OTmZWbNmcfToUeLj47n++usJCQkp7bKEEEKIIlP+v62vlsvhnXB9ZC30fhm3tSb/WnKI/aeyeXdoC6pZ/V/UVGvN6dzTjFs1jlO5p5h4zURqhdXCYrQUX/0lJCcnh2nTpmGxWLjpppvO3XEmhBBCVCQSjDKTYckEqNYEWvyNXac9/Ov3/fRoWIXrmlbHVMB5RR9v+pgNJzfwTLtnaFGlBSGm8ntGRWvNgQMHSEhIIDg4mKFDh1K7dm0CA/0Pi0IIIUR5UrnbELsc8Pv7kH0Seo3BYY7gzbk7AXjmukRCzAWbVzRz70z+u/u/3FT/JgYkDMBqKb+TrVNSUvjiiy/48ssv2b9/PwANGzaUUCSEEKJCq9xnjI5tgA1fQtNboE4nZm1LYemfp/hHz/okVg/z+1KRy+Ni7fG1TFw/kVZVW/Fwy4cJt4RjUOUvdzocDpYuXcrKlSsxm80MHDiQ+Pj40i5LCCGEKBGVNxh53LDkLTAFQ/eRpDotTPh1HXVjQri3a7zfl9C01uxL28eY5WOIDoxmTEfvf8vjZGutNZ9//jnJycm0bNmS3r17y+RqIYQQlUr5+/YuKsmbYO9C6PgwOjKej+bsITndxqd3tiEy2Oz3blJtqYxeMZpsZzYf9vqQWtZa5a6zdVpaGlarFYPBQPfu3QkODqZ27dqlXZYQQghR4ipnMNIaVn0MRgu0uYfNx218veoQA5rF0qNhVb8voeU6c5mwdgLbU7czpuMYWlRpUa7uQHO5XKxYsYJly5bRp08f2rdvT1JSUmmXJYQQQpSayhmMzhyC7f+DJjfhDI9n3A9rCTYbeaZfIuYA/y6huTwuvt7xNXP2z2FEoxH0T+hfrpb72Lt3L7Nnz+b06dM0btxYApEQQghBMd+VppTqp5TapZTao5R6Lp/Xuyul/lBKuZRStxRnLRdYOwXcTuj4EN//cYz1h9L4R6/61InyL9horVl6ZCmTN06mY2xHHmzxIKHm8rOC/IIFC/j6668BGDFiBEOHDsVqLb930AkhhBBFpdjOGCmljMBkoA9wBFirlJqptd6eZ7NDwN3AyOKq4y9yTnuX/qh3LSdDGvLuvOU0ibNyZ8d4vy+h7U7bzZjlY4gLjWNsx7GEm8OLueir5/F4cLvdmEwm6tWrh8lkokuXLgQEVM6ThkIIIUR+ivNbsT2wR2u9D0Ap9W9gMHAuGGmtD/he8xRjHRf64wvvEiAdH2LCvD2k5zr55I42BPrZsyjNlsaopaNwaRevd32dGmE1ynwH6EOHDjF79mwSEhK47rrrSEhIICEhobTLEkIIIcqc4gxGNYDDeR4fAToUZkdKqQeAB4Cru1vKZYe1n0FsS1apZvx3/TqGtatF+/gov97udDsZs2IMe9P28nq312kWU7aXxcjOzmbBggVs3LgRq9VKnTp1SrskIYQQokwrzmCUX2LQhdmR1noqMBWgbdu2hdoH4F0TLf0I7mte4uVfdhETauGZ6xL9Cjdaa6ZsnsLiw4v5e7O/0y++H0aD/52xS9ru3buZMWMGDoeDLl260L17d8xm/9sQCCGEEJVRcQajI0CtPI9rAseK8XiX5/HAqn9CRB2+yGjFzuN7mXBLc6JD/bu9fuGhhUzdPJWetXryYPMHy2wDR601SimioqKIi4ujX79+VKlSpbTLEkIIIcqF4rwrbS3QQCmVoJQyA8OBmcV4vMvbtxiOb8He6l4mLT1Mq9oRDGlVw6+3Hss8xisrXyHeGs+4zuMwB5S9My82m405c+Ywffp0AKKjo7njjjskFAkhhBAFUGzBSGvtAv4B/ArsAP6jtd6mlBqnlBoEoJRqp5Q6AgwFpiilthVXPaz4CIIi+VduD05nO3imbyJGP5b9cLldPPf7c9hcNt7o9gYRgRHFVmJhaK3ZsmULkydPZs2aNQQHB+PxlNxcdiGEEKIiKdbrQVrr2cDsi54bk+fntXgvsRWv41tg32Js7f/BJ6tO0rV+DJ3rx/j11smbJrPh5Aaeb/88TWKaFHOhBZOens5PP/3E/v37iYuL47bbbiMuLq60yxJCCCHKrbI5UaaoLf8QjBamOvuRYctg5HUN/XrbymMrmbZlGr1r92Z44vBiLrLgTCYT6enpDBgwgNatW2MwFGu/TiGEEKLCq/jBKOMYbP+R3KQhfLwum76Nq9GyVuQV33Y69zQv/P4CsaGxvNz55TIROrTW7Nq1i02bNjF06FCCg4N55JFHykRtQgghREVQ8YPRysngdjHVfQN2l5uRfROv+BatNc///jzp9nT+33X/j3BL6Xe2PnPmDHPnzuXPP/+katWqZGVlYbVaJRQJIYQQRahiByN7Fmz4itz4a5m8xcigFtVpWD3sim+btnUaK46t4Mk2T9KiaosSKPTS3G43y5cvZ9myZSil6NOnDx06dMBoLLs9lIQQQojyqmIHow1fgS2daQzCozVP97ny2aJNKZuYtGESXeO6ck+Te0qgyMvTWrN582YaNmzIddddJ4u9CiGEEMWo4gYjjwdWT8Ee04T3dsUwvH0takUHX/YtWY4sRi0dRWRgJOO7ji+15T4yMzNZunQpffr0wWw2c//99xMYGFgqtQghhBCVScUNRn/+Cmf282215wkwGnj82gaX3VxrzZgVY0jOTubTvp8SFeTf+mlFyePxsGbNGhYvXozb7SYpKYl69epJKBJCCCFKSMUNRqsm4QyqyusHG3F3tzpUtV4+XEz/czrzD87nwRYP0r56+xIq8rzDhw8za9YsTpw4Qf369enfvz9RUSUfzoQQQojKrGIGoxPb4MDvLKx6P8ZMM4/0rH/ZzXef2c1ba9+ibbW2PNj8wRIq8kK//fYbubm53HrrrSQlJZXaZTwhhBCiMquYwWjFJHRAIONOdOa6ptWJCL702ma5rlyeXvI0wQHBTOg+AaOhZO720lqzYcMG6tevj9VqZfDgwQQGBmI2l7112IQQQojKouI1wclKga3T2R83kGP2QIa3q33Zzd9Y/Qb70/czvut4qgSXzIKrx48fZ9q0afz888+sX78eAKvVKqFICCGEKGUV74zR2s/A7WBS7nXUjAyiQ8Kl5+nM3jebH/f8yF1N7qJbzW7FXprNZmPx4sWsXbuW4OBgbrzxRpo3b17sxxVCCCGEfypWMHLZYd2/yK3Vgxm7Q3iydy0Mhvzn6hzOPMwrK1+hWUwzHm/9eImU99tvv7FmzRratm1Lr169CAoKKpHjCiGEEMI/FSsY7ZoD2SnMqn0zChjatma+mzncDkb+NhKDMvB2j7cxGUzFVtKpU6fweDxUrVqV7t2707x5c+Li4orteEIIIYQovIoVjHb8jA6KYuLeOLo2CCcuIv8zMuNXjWf76e1MvGYiNUJrFEspTqeTpUuXsmLFCuLj47njjjsIDg4mOPjyTSaFEEIIUXoqTjByO2H3PE7E9eboDicvDMx/0vV/dv2HGXtmcE+Te+hdp3exlLJr1y7mzJlDeno6LVq0oE+fPsVyHCGEEEIUrYoTjA78DvYM/mdrTXiQid6Nq/5lk00nN/HmmjfpUL1Dsc0r2rJlCzNmzKBKlSrcfffd1KlTp1iOI4QQQoiiV3GC0c5f0AFBTDpUiyHta2AJuLAf0ancUzz525PEBMXwTo93irRfkcvlIi0tjZiYGBo1asSAAQNo1aoVRmPJ9EQSQgghRNGoGMFIa9g5i0ORHcnKMjGsXa0LXnZ6nDz121OkO9L5qt9XRARGFNmh9+3bx+zZs3E6nTz66KMEBATQtm3bItu/EEIIIUpOxQhGyRshM5np7uE0jrPSOM56wcsT1k5gw8kNvNblNRrHNC6SQ2ZmZjJv3jy2bt1KZGQkN9xwAwEBFWM4hRBCiMqqYnyT7/gFrQx8dTqJJwddeIv+zD0z+ffOfzMscRiD6w8uksOlpqYydepU3G43PXr0oGvXrhKKhBBCiAqgYnyb7/yFAyEtyHGEM7jl+dvvt5/azrhV42hZpSWj2o+66sNkZ2cTEhJCVFQU7du3p2XLlkRHR1/1foUQQghRNpT/tdJS90LKTjaFdqV/s/MLxuY4cxi5dCRh5jAmXjPxqpo45uTkMHPmTD766CMyMjJQSnHttddKKBJCCCEqmPJ/xmjnLABuHPZ3BoWf7130+urXOZJ5hKl9phITHFOoXWut2bBhAwsWLMBms9GxY0cCAwOLpGwhhBBClD0VIxhVawqRdc6d/pq7fy4/7f2JuxrfRce4joXarcvl4ssvv+Tw4cPUrl2bAQMGULXqX3sjCSGEEKLiKN/ByOOBmAZQtdG5p5Kzkhm3ahxJUUmFauLodrsxGo0EBARQs2ZN2rRpQ/PmzVEq/8VohRBCCFFxlO9gZDDA4EnnHmqteWn5SzjdTiZ0n4DJ6P+8Iq0127ZtY/78+QwfPpzY2Fj69u1bHFULIYQQoowq38HoIvMPzmfN8TWMajeKhPAEv96jXS5STp1i7rx57N+/n9jYWDk7JIQQQlRSFSYY2d123ln3DnXD6zI8afgVt9da405PZ/5nn7EuMxNTUBDXX389bdq0wWAo/zfrCSGEEKLgKkww+nzr5yRnJzO1z1QCDJf/WO7cXLIWLCDlo0lkW8NIiI9nwN33EJkQXyK1CiGEEKJsqhDBKCUnhX9t/Rfda3anU1ynS26nXS6SV65k7syfqf3HH8QrxbW3/Y3wGwZiMJtLsGIhhBBClEUVIhhNXD8Rl8fFqHaX7m6de/AQiz6dygalINBCvf79ib/vXozh4Si5dCaEEEIIKkAw2npqKz/v+5m7mtxFbWvtv7zucbnY9sUXLNi5k4zQUOLdbgYMv43oxIYoo7EUKhZCCCFEWVXug9H21O1UC67Gg80f/Mtr9r17Of7KOI6cPImndSuGtGtHk759UbLgqxBCCCHyobTWpV1DgbRt21avW7fugudsLhuBAeeX6nBmZrJk8j+xr1pFvYMHiX7gAay3j8ASFlbS5QohhBBXRSm1XmvdtrTrqCwqxKmTs6FIezzs+PprFmzezJmwMOITG9L73Xew1KtXyhUKIYQQojyoEMEI4NTvy/n1P9+zJzqaYIuFgY0b0+rmmzHIPCIhhBBC+KncByOPzcbJtybw54IF7O3Vk9bR0fS+916CgoNLuzQhhBBClDPFGoyUUv2ADwAj8JnW+s2LXrcAXwJtgFRgmNb6gL/7P7R2LVsn/5OEVatoMHw4zR9+iIiqVYvuAwghhBCiUim2YKSUMgKTgT7AEWCtUmqm1np7ns3uA85oresrpYYDbwHDrrTv3MxMfp0yhc2ZmVhiY+kw6SOie/cujo8hhBBCiEqkOM8YtQf2aK33ASil/g0MBvIGo8HAy76fpwOTlFJKX+ZWuZy0ND56801yTSYSs7Lo9+CDRDRoUDyfQAghhBCVSnEGoxrA4TyPjwAdLrWN1tqllEoHooFTl9ppek4OgTYbN3XqRIOBA4u4ZCGEEEJUZsUZjFQ+z118JsifbVBKPQA84Htof/y997Y+/t57V1lepRPDZQKnuCQZt4KTMSscGbfCqQzjVqe0C6hMijMYHQFq5XlcEzh2iW2OKKUCgHDg9MU70lpPBaYCKKXWSaOrgpNxKxwZt4KTMSscGbfCkXETRa04V09dCzRQSiUopczAcGDmRdvMBO7y/XwLsOhy84uEEEIIIYpTsZ0x8s0Z+gfwK97b9adprbcppcYB67TWM4F/AV8ppfbgPVM0vLjqEUIIIYS4kmLtY6S1ng3Mvui5MXl+tgFDC7jbqUVQWmUk41Y4Mm4FJ2NWODJuhSPjJopUuVtEVgghhBCiuBTnHCMhhBBCiHJFgpEQQgghhE+ZDUZKqX5KqV1KqT1Kqefyed2ilPre9/pqpVR8yVdZtvgxZt2VUn8opVxKqVtKo8ayyI9xe0optV0ptVkptVApJT1F8GvcHlRKbVFKbVRK/a6UalwadZY1Vxq3PNvdopTSSqlKfyu6H79rdyulUny/axuVUveXRp2iYiiTwSjPOmv9gcbAbfn8pXpunTVgIt511iotP8fsEHA38G3JVld2+TluG4C2WuvmeJeumVCyVZY9fo7bt1rrZlrrlnjHrNJ3ZfVz3FBKhQGPAatLtsKyx98xA77XWrf0/fmsRIsUFUqZDEbkWWdNa+0Azq6zltdg4Avfz9OBa5VS+XXSriyuOGZa6wNa682ApzQKLKP8GbfFWusc38NVeJuVVnb+jFtGnoch5NPVvhLy5+82gFfxhklbSRZXRvk7ZkIUibIajPJbZ63GpbbRWruAs+usVVb+jJn4q4KO233AnGKtqHzwa9yUUo8opfbi/ZJ/rIRqK8uuOG5KqVZALa31LyVZWBnm7/+jQ3yXu6crpWrl87oQfimrwajI1lmrRGQ8CsfvcVNK3Q60Bd4u1orKB7/GTWs9WWtdDxgFvFTsVZV9lx03pZQB79SAp0usorLPn9+1n4F43+XuBZy/miBEgZXVYFSQdda43DprlYg/Yyb+yq9xU0r1Bl4EBmmt7SVUW1lW0N+3fwM3FmtF5cOVxi0MaAr8ppQ6AHQEZlbyCdhX/F3TWqfm+f/yU6BNCdUmKqCyGoxknbWC82fMxF9dcdx8lzam4A1FJ0uhxrLIn3FrkOfhAGB3CdZXVl123LTW6VrrGK11vNY6Hu+ctkFa63WlU26Z4M/vWmyeh4OAHSVYn6hginVJkMKSddYKzp8xU0q1A34EIoEblFKvaK2blGLZpc7P37W3gVDgB9/8/kNa60GlVnQZ4Oe4/cN3ps0JnOH8P2QqLT/HTeTh55g9ppQaBLjwfh/cXWoFi3JPlgQRQgghhPApq5fShBBCCCFKnAQjIYQQQggfCUZCCCGEED4SjIQQQgghfCQYCSGEEEL4SDASooQppdx5VgHfqJSKv8y28UqprUVwzN98q5NvUkotV0olFmIfDyql7vT9fLdSKi7Pa59dYmHPq6lzrVKqpR/veUIpFXy1xxZCCJBgJERpyM2zCnhLrfWBEjruCK11C7zLJRR4WROt9Sda6y99D+8G4vK8dr/WenuRVHm+zn/iX51PABKMhBBFQoKREGWA78zQMqXUH74/nfPZpolSao3vLNPms52llVK353l+ilLKeIXDLQXq+957rVJqg1Jqi1JqmlLK4nv+TaXUdt9x3vE997JSaqRS6ha8a8Z94ztmkO9MT1ul1ENKqQl5ar5bKfVRIetcSZ7FQpVSHyul1imltimlXvE99xjegLZYKbXY91xfpdRK3zj+oJQKvcJxhBDiHAlGQpS8oDyX0X70PXcS6KO1bg0MAz7M530PAh9orVviDSZHlFKNfNt38T3vBkZc4fg3AFuUUoHA58AwrXUzvJ3wH1JKRQE3AU18i3K+lvfNWuvpwDq8Z3Zaaq1z87w8Hbg5z+NhwPeFrLMf8L88j1/UWrcFmgM9lFLNtdYf4l03q6fWuqdSKgbvYrW9fWO5DnjqCscRQohzyuSSIEJUcLm+cJCXCZjkm1PjBhrm876VwItKqZrADK31bqXUtXgXzFzrW64kCG/Iys83Sqlc4ADwKJAI7Nda/+l7/QvgEWASYAM+U0rNAn7x94NprVOUUvuUUh3xro2WCCz37bcgdYbgXf6hdZ7nb1VKPYD3761YoDGw+aL3dvQ9v9x3HDPecRNCCL9IMBKibHgSOAG0wHsm13bxBlrrb5VSq/EuyPqrUup+QAFfaK2f9+MYI/IuRqqUis5vI9/aVO2Ba/GuQfgPoFcBPsv3wK3ATuBHrbVW3pTid53AJuBNYDJws1IqARgJtNNan1FKfQ4E5vNeBczXWt9WgHqFEOIcuZQmRNkQDiRrrT3AHXjPllxAKVUX2Oe7fDQT7yWlhcAtSqmqvm2ilFJ1/DzmTiBeKVXf9/gOYIlvTk641no23onN+d0ZlgmEXWK/M4AbgdvwhiQKWqfW2on3klhH32U4K5ANpCulqgH9L1HLKqDL2c+klApWSuV39k0IIfIlwUiIsuGfwF1KqVV4L6Nl57PNMGCrUmojkAR86bsT7CVgnlJqMzAf72WmK9Ja24B7gB+UUlsAD/AJ3pDxi29/S/CezbrY58AnZydfX7TfM8B2oI7Weo3vuQLX6Zu79C4wUmu9CdgAbAOm4b08d9ZUYI5SarHWOgXvHXPf+Y6zCu9YCSGEX5TWurRrEEIIIYQoE+SMkRBCCCGEjwQjIYQQQggfCUZCCCGEED4SjIQQQgghfCQYCSGEEEL4SDASQgghhPCRYCSEEEII4fP/AYMMpiPjlO3gAAAAAElFTkSuQmCC\n", "text/plain": [ "
" ] @@ -540,12 +540,12 @@ }, { "cell_type": "code", - "execution_count": 24, + "execution_count": 23, "metadata": {}, "outputs": [ { "data": { - "image/png": "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\n", + "image/png": "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\n", "text/plain": [ "
" ] @@ -560,7 +560,7 @@ }, { "cell_type": "code", - "execution_count": 25, + "execution_count": 24, "metadata": {}, "outputs": [ { @@ -603,10 +603,10 @@ " PureSVD\n", " CoFFee\n", " CoFFee\n", - " mostpopular\n", - " mostpopular\n", - " random\n", - " random\n", + " MP\n", + " MP\n", + " RND\n", + " RND\n", " \n", " \n", " top-n\n", @@ -625,160 +625,160 @@ " 1\n", " 0.077220\n", " 0.026973\n", - " 0.069769\n", - " 0.019528\n", + " 0.066921\n", + " 0.018923\n", " 0.034049\n", " 0.021976\n", - " 0.000390\n", - " 0.001630\n", + " 0.000627\n", + " 0.001624\n", " \n", " \n", " 2\n", " 0.111342\n", " 0.040239\n", - " 0.098721\n", - " 0.029681\n", + " 0.095919\n", + " 0.027686\n", " 0.049284\n", " 0.031109\n", - " 0.000552\n", - " 0.002211\n", + " 0.000843\n", + " 0.002143\n", " \n", " \n", " 3\n", " 0.132601\n", " 0.051971\n", - " 0.117231\n", - " 0.037676\n", + " 0.114597\n", + " 0.036112\n", " 0.057795\n", " 0.037549\n", - " 0.000748\n", - " 0.002522\n", + " 0.000955\n", + " 0.002853\n", " \n", " \n", " 5\n", " 0.159641\n", " 0.069481\n", - " 0.142215\n", - " 0.049924\n", + " 0.139822\n", + " 0.048029\n", " 0.070335\n", " 0.048499\n", - " 0.001120\n", - " 0.002918\n", + " 0.001335\n", + " 0.003335\n", " \n", " \n", " 10\n", " 0.197528\n", " 0.101178\n", - " 0.178211\n", - " 0.072297\n", + " 0.175671\n", + " 0.070128\n", " 0.091356\n", " 0.064922\n", - " 0.001662\n", - " 0.004274\n", + " 0.002206\n", + " 0.004683\n", " \n", " \n", " 15\n", " 0.219348\n", " 0.122217\n", - " 0.199277\n", - " 0.088928\n", + " 0.195807\n", + " 0.087583\n", " 0.105611\n", " 0.077387\n", - " 0.002112\n", - " 0.005217\n", + " 0.002830\n", + " 0.006075\n", " \n", " \n", " 20\n", " 0.234971\n", " 0.139544\n", - " 0.213987\n", - " 0.101211\n", + " 0.211216\n", + " 0.100503\n", " 0.116623\n", " 0.086674\n", - " 0.002620\n", - " 0.006157\n", + " 0.003368\n", + " 0.006831\n", " \n", " \n", " 30\n", " 0.256997\n", " 0.166981\n", - " 0.235988\n", - " 0.123764\n", + " 0.232862\n", + " 0.122439\n", " 0.130844\n", " 0.101282\n", - " 0.003894\n", - " 0.007529\n", + " 0.004546\n", + " 0.008924\n", " \n", " \n", " 50\n", " 0.284766\n", " 0.205027\n", - " 0.263548\n", - " 0.154949\n", + " 0.260017\n", + " 0.152073\n", " 0.151501\n", " 0.125126\n", - " 0.005715\n", - " 0.010865\n", + " 0.006236\n", + " 0.012502\n", " \n", " \n", " 70\n", " 0.302296\n", " 0.230614\n", - " 0.280526\n", - " 0.179861\n", + " 0.277340\n", + " 0.176985\n", " 0.164905\n", " 0.143222\n", - " 0.007435\n", - " 0.013285\n", + " 0.007707\n", + " 0.015067\n", " \n", " \n", " 100\n", " 0.319825\n", " 0.262407\n", - " 0.297796\n", - " 0.205782\n", + " 0.294639\n", + " 0.202910\n", " 0.180364\n", " 0.165874\n", - " 0.009852\n", - " 0.017074\n", + " 0.009602\n", + " 0.019054\n", " \n", " \n", "\n", "" ], "text/plain": [ - " nDCG nDCL nDCG nDCL nDCG nDCL \\\n", - " PureSVD PureSVD CoFFee CoFFee mostpopular mostpopular \n", - "top-n \n", - "1 0.077220 0.026973 0.069769 0.019528 0.034049 0.021976 \n", - "2 0.111342 0.040239 0.098721 0.029681 0.049284 0.031109 \n", - "3 0.132601 0.051971 0.117231 0.037676 0.057795 0.037549 \n", - "5 0.159641 0.069481 0.142215 0.049924 0.070335 0.048499 \n", - "10 0.197528 0.101178 0.178211 0.072297 0.091356 0.064922 \n", - "15 0.219348 0.122217 0.199277 0.088928 0.105611 0.077387 \n", - "20 0.234971 0.139544 0.213987 0.101211 0.116623 0.086674 \n", - "30 0.256997 0.166981 0.235988 0.123764 0.130844 0.101282 \n", - "50 0.284766 0.205027 0.263548 0.154949 0.151501 0.125126 \n", - "70 0.302296 0.230614 0.280526 0.179861 0.164905 0.143222 \n", - "100 0.319825 0.262407 0.297796 0.205782 0.180364 0.165874 \n", + " nDCG nDCL nDCG nDCL nDCG nDCL nDCG \\\n", + " PureSVD PureSVD CoFFee CoFFee MP MP RND \n", + "top-n \n", + "1 0.077220 0.026973 0.066921 0.018923 0.034049 0.021976 0.000627 \n", + "2 0.111342 0.040239 0.095919 0.027686 0.049284 0.031109 0.000843 \n", + "3 0.132601 0.051971 0.114597 0.036112 0.057795 0.037549 0.000955 \n", + "5 0.159641 0.069481 0.139822 0.048029 0.070335 0.048499 0.001335 \n", + "10 0.197528 0.101178 0.175671 0.070128 0.091356 0.064922 0.002206 \n", + "15 0.219348 0.122217 0.195807 0.087583 0.105611 0.077387 0.002830 \n", + "20 0.234971 0.139544 0.211216 0.100503 0.116623 0.086674 0.003368 \n", + "30 0.256997 0.166981 0.232862 0.122439 0.130844 0.101282 0.004546 \n", + "50 0.284766 0.205027 0.260017 0.152073 0.151501 0.125126 0.006236 \n", + "70 0.302296 0.230614 0.277340 0.176985 0.164905 0.143222 0.007707 \n", + "100 0.319825 0.262407 0.294639 0.202910 0.180364 0.165874 0.009602 \n", "\n", - " nDCG nDCL \n", - " random random \n", - "top-n \n", - "1 0.000390 0.001630 \n", - "2 0.000552 0.002211 \n", - "3 0.000748 0.002522 \n", - "5 0.001120 0.002918 \n", - "10 0.001662 0.004274 \n", - "15 0.002112 0.005217 \n", - "20 0.002620 0.006157 \n", - "30 0.003894 0.007529 \n", - "50 0.005715 0.010865 \n", - "70 0.007435 0.013285 \n", - "100 0.009852 0.017074 " + " nDCL \n", + " RND \n", + "top-n \n", + "1 0.001624 \n", + "2 0.002143 \n", + "3 0.002853 \n", + "5 0.003335 \n", + "10 0.004683 \n", + "15 0.006075 \n", + "20 0.006831 \n", + "30 0.008924 \n", + "50 0.012502 \n", + "70 0.015067 \n", + "100 0.019054 " ] }, - "execution_count": 25, + "execution_count": 24, "metadata": {}, "output_type": "execute_result" } @@ -789,12 +789,12 @@ }, { "cell_type": "code", - "execution_count": 26, + "execution_count": 25, "metadata": {}, "outputs": [ { "data": { - "image/png": "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\n", + "image/png": "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\n", "text/plain": [ "
" ] From d624ac7fba394abf1025c9fa97b5c867b3d976ac Mon Sep 17 00:00:00 2001 From: Evgeny Frolov Date: Sun, 11 Mar 2018 14:44:55 +0300 Subject: [PATCH 21/21] update release version --- setup.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/setup.py b/setup.py index eb36ce4..6563e77 100644 --- a/setup.py +++ b/setup.py @@ -17,7 +17,7 @@ opts = dict(name="polara", description="Fast and flexible recommender framework", keywords = "recommender system", - version = "0.5.1", + version = "0.5.2", license="MIT", author="Evgeny Frolov", platforms=["any"],