Dimen_reduc.ipynb

{"nbformat":4,"nbformat_minor":0,"metadata":{"colab":{"name":"Daniel Copy of 254FinalLab.ipynb","provenance":[{"file_id":"1vmdRQ0s0s1W37woreWhAiWvO-_pNLKje","timestamp":1576207500073}],"collapsed_sections":[]},"kernelspec":{"name":"python3","display_name":"Python 3"}},"cells":[{"cell_type":"code","metadata":{"id":"d7JCQFifFq47","colab_type":"code","outputId":"a2d75089-cd7e-49a9-f8d0-daf7cbcd0d65","executionInfo":{"status":"ok","timestamp":1576268935426,"user_tz":480,"elapsed":94406,"user":{"displayName":"Daniel McAndrew","photoUrl":"","userId":"01097539478049760558"}},"colab":{"base_uri":"https://localhost:8080/","height":199}},"source":["import numpy as np\n","import pandas as pd\n","import matplotlib.pyplot as plt\n","\n","from sklearn.manifold import TSNE\n","from sklearn.metrics import accuracy_score\n","from sklearn.metrics import mean_absolute_error\n","from gensim.test.utils import common_texts\n","from gensim.models.doc2vec import Doc2Vec, TaggedDocument\n","\n","import nltk\n","nltk.download('stopwords')\n","nltk.download('punkt')\n","from nltk.corpus import stopwords \n","from nltk.tokenize import word_tokenize \n","stop_words = set(stopwords.words('english'))\n","from google.colab import drive\n","drive.mount('/content/drive') "],"execution_count":0,"outputs":[{"output_type":"stream","text":["[nltk_data] Downloading package stopwords to /root/nltk_data...\n","[nltk_data]   Unzipping corpora/stopwords.zip.\n","[nltk_data] Downloading package punkt to /root/nltk_data...\n","[nltk_data]   Unzipping tokenizers/punkt.zip.\n","Go to this URL in a browser: https://accounts.google.com/o/oauth2/auth?client_id=947318989803-6bn6qk8qdgf4n4g3pfee6491hc0brc4i.apps.googleusercontent.com&redirect_uri=urn%3aietf%3awg%3aoauth%3a2.0%3aoob&response_type=code&scope=email%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdocs.test%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdrive%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdrive.photos.readonly%20https%3a%2f%2fwww.googleapis.com%2fauth%2fpeopleapi.readonly\n","\n","Enter your authorization code:\n","··········\n","Mounted at /content/drive\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"kxQhLw-XFtkX","colab_type":"code","outputId":"fb6486ef-bdb3-436b-8ac0-c0cd085faa02","executionInfo":{"status":"ok","timestamp":1576268937475,"user_tz":480,"elapsed":96445,"user":{"displayName":"Daniel McAndrew","photoUrl":"","userId":"01097539478049760558"}},"colab":{"base_uri":"https://localhost:8080/","height":408}},"source":["# Read in the wine review data:\n","df = pd.read_csv(\"./drive/My Drive/winemag-data-130k-v2.csv\", index_col=0)\n","df.head()"],"execution_count":0,"outputs":[{"output_type":"execute_result","data":{"text/html":["<div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>country</th>\n","      <th>description</th>\n","      <th>designation</th>\n","      <th>points</th>\n","      <th>price</th>\n","      <th>province</th>\n","      <th>region_1</th>\n","      <th>region_2</th>\n","      <th>taster_name</th>\n","      <th>taster_twitter_handle</th>\n","      <th>title</th>\n","      <th>variety</th>\n","      <th>winery</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>Italy</td>\n","      <td>Aromas include tropical fruit, broom, brimston...</td>\n","      <td>Vulkà Bianco</td>\n","      <td>87</td>\n","      <td>NaN</td>\n","      <td>Sicily &amp; Sardinia</td>\n","      <td>Etna</td>\n","      <td>NaN</td>\n","      <td>Kerin O’Keefe</td>\n","      <td>@kerinokeefe</td>\n","      <td>Nicosia 2013 Vulkà Bianco  (Etna)</td>\n","      <td>White Blend</td>\n","      <td>Nicosia</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>Portugal</td>\n","      <td>This is ripe and fruity, a wine that is smooth...</td>\n","      <td>Avidagos</td>\n","      <td>87</td>\n","      <td>15.0</td>\n","      <td>Douro</td>\n","      <td>NaN</td>\n","      <td>NaN</td>\n","      <td>Roger Voss</td>\n","      <td>@vossroger</td>\n","      <td>Quinta dos Avidagos 2011 Avidagos Red (Douro)</td>\n","      <td>Portuguese Red</td>\n","      <td>Quinta dos Avidagos</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>US</td>\n","      <td>Tart and snappy, the flavors of lime flesh and...</td>\n","      <td>NaN</td>\n","      <td>87</td>\n","      <td>14.0</td>\n","      <td>Oregon</td>\n","      <td>Willamette Valley</td>\n","      <td>Willamette Valley</td>\n","      <td>Paul Gregutt</td>\n","      <td>@paulgwine</td>\n","      <td>Rainstorm 2013 Pinot Gris (Willamette Valley)</td>\n","      <td>Pinot Gris</td>\n","      <td>Rainstorm</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>US</td>\n","      <td>Pineapple rind, lemon pith and orange blossom ...</td>\n","      <td>Reserve Late Harvest</td>\n","      <td>87</td>\n","      <td>13.0</td>\n","      <td>Michigan</td>\n","      <td>Lake Michigan Shore</td>\n","      <td>NaN</td>\n","      <td>Alexander Peartree</td>\n","      <td>NaN</td>\n","      <td>St. Julian 2013 Reserve Late Harvest Riesling ...</td>\n","      <td>Riesling</td>\n","      <td>St. Julian</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>US</td>\n","      <td>Much like the regular bottling from 2012, this...</td>\n","      <td>Vintner's Reserve Wild Child Block</td>\n","      <td>87</td>\n","      <td>65.0</td>\n","      <td>Oregon</td>\n","      <td>Willamette Valley</td>\n","      <td>Willamette Valley</td>\n","      <td>Paul Gregutt</td>\n","      <td>@paulgwine</td>\n","      <td>Sweet Cheeks 2012 Vintner's Reserve Wild Child...</td>\n","      <td>Pinot Noir</td>\n","      <td>Sweet Cheeks</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>"],"text/plain":["    country  ...               winery\n","0     Italy  ...              Nicosia\n","1  Portugal  ...  Quinta dos Avidagos\n","2        US  ...            Rainstorm\n","3        US  ...           St. Julian\n","4        US  ...         Sweet Cheeks\n","\n","[5 rows x 13 columns]"]},"metadata":{"tags":[]},"execution_count":3}]},{"cell_type":"code","metadata":{"id":"G3dkvkgNFydf","colab_type":"code","colab":{}},"source":["# Create a tokenized set of documents which are lists of words included in each description:\n","docs = list(df[\"description\"])\n","# use this regex to remove punctuation\n","pattern = r'''(?x)  # set flag to allow verbose regexps\n","(?:[A-Z]\\.)+        # abbreviations, e.g. U.S.A.\n","|\\w+(?:[-']\\w+)*    # words with optional internal hyphens\n","|\\$?\\d+(?:\\.\\d+)?   # currency, e.g. $12.80 \n","|\\.\\.\\.             # elipses\n","|[.,;\"'?()-_`]      # these are separate tokens\n","'''\n","# Iterate through the docs and tokenize them\n","for i in range(len(docs)):\n","  word_tokens = word_tokenize(\" \".join(nltk.regexp_tokenize(docs[i], pattern))) \n","  docs[i] = filtered_sentence = [w.lower() for w in word_tokens if not w.lower() in stop_words]"],"execution_count":0,"outputs":[]},{"cell_type":"code","metadata":{"id":"VrHDgY5DF2fv","colab_type":"code","colab":{}},"source":["# Create a list of taggedDocument objects for use in the doc2vec model\n","documents = [TaggedDocument(doc, [str(i)]) for i, doc in enumerate(docs)]"],"execution_count":0,"outputs":[]},{"cell_type":"code","metadata":{"id":"-BtAAfwBF3cP","colab_type":"code","colab":{}},"source":["# Train a doc2vec model on our descriptions:\n","model = Doc2Vec(documents, vector_size=10, window=10, min_count=3, workers=4)"],"execution_count":0,"outputs":[]},{"cell_type":"code","metadata":{"id":"04OsK3vfF6VH","colab_type":"code","colab":{}},"source":["# Define a function to create a document vector for each wine using our trained model:\n","def get_doc_vec(doc):\n","  return model.infer_vector(doc.split())"],"execution_count":0,"outputs":[]},{"cell_type":"code","metadata":{"id":"fi06VHE2F9ZW","colab_type":"code","colab":{}},"source":["# Create index feature for use in the closest document function\n","df[\"idx\"] = np.arange(len(df))\n","# apply the document vector function:\n","df[\"document_vector\"] = df[\"description\"].apply(get_doc_vec)"],"execution_count":0,"outputs":[]},{"cell_type":"code","metadata":{"id":"-L39y95dF_an","colab_type":"code","colab":{}},"source":["#df.head()\n","#df.to_csv(\"./drive/My Drive/winemag-data-130k-v2_docvecs.csv\", index=False)"],"execution_count":0,"outputs":[]},{"cell_type":"code","metadata":{"id":"PgOKPPcsnhYH","colab_type":"code","colab":{}},"source":["# Find the 5 most similar wines by deescription:\n","def closest_descriptions(idx):\n","  sims = model.docvecs.most_similar([idx], topn=5)\n","  return [sim[0] for sim in sims]"],"execution_count":0,"outputs":[]},{"cell_type":"code","metadata":{"id":"UM4Wdeiq8bgY","colab_type":"code","outputId":"d6e989a0-a59b-40a9-a708-28d9776ebad7","executionInfo":{"status":"ok","timestamp":1576269778095,"user_tz":480,"elapsed":437781,"user":{"displayName":"Daniel McAndrew","photoUrl":"","userId":"01097539478049760558"}},"colab":{"base_uri":"https://localhost:8080/","height":73}},"source":["# Apply the closest descriptions function:\n","df[\"closest_description\"] = df[\"idx\"].apply(closest_descriptions)"],"execution_count":0,"outputs":[{"output_type":"stream","text":["/usr/local/lib/python3.6/dist-packages/gensim/matutils.py:737: FutureWarning: Conversion of the second argument of issubdtype from `int` to `np.signedinteger` is deprecated. In future, it will be treated as `np.int64 == np.dtype(int).type`.\n","  if np.issubdtype(vec.dtype, np.int):\n"],"name":"stderr"}]},{"cell_type":"code","metadata":{"id":"06zaoUti_VK9","colab_type":"code","outputId":"ad3b34cd-3a3b-4913-9afe-c0b47e653e3c","executionInfo":{"status":"ok","timestamp":1576270164814,"user_tz":480,"elapsed":397,"user":{"displayName":"Daniel McAndrew","photoUrl":"","userId":"01097539478049760558"}},"colab":{"base_uri":"https://localhost:8080/","height":523}},"source":["# Inspect some results:\n","for i in range(3):\n","  print(i, \"\\n\", df[\"description\"][i], \"\\n\")\n","  for j in range(2):\n","    print(df[\"description\"].iloc[int(df[\"closest_description\"][i][j])], \"\\n\\n\")"],"execution_count":0,"outputs":[{"output_type":"stream","text":["0 \n"," Aromas include tropical fruit, broom, brimstone and dried herb. The palate isn't overly expressive, offering unripened apple, citrus and dried sage alongside brisk acidity. \n","\n","A beautifully fragrant Grillo, this has a flower aroma and an evident touch of zesty spice on the finish. The mouthfeel is lean and fresh. \n","\n","\n","You'll find bright berry notes followed by white licorice, forest floor and a distant touch of chopped herbs in this fresh, well-integrated Barbera from Monferrato. The mouthfeel is not too concentrated and the wine is naturally acidic (try it with cheese- or butter-based pasta dishes). \n","\n","\n","1 \n"," This is ripe and fruity, a wine that is smooth while still structured. Firm tannins are filled out with juicy red berry fruits and freshened with acidity. It's  already drinkable, although it will certainly be better from 2016. \n","\n","Bouncy berry and cherry aromas are more or less standard for basic Garnacha. This wine feels a bit gummy and grabby, but not sticky. Lightly herbal, minty dark-berry flavors finish herbal, chocolaty and a touch green. \n","\n","\n","Medium sweet in style, this is as much about ripe tropical and pineapple fruits as about spice and richness. It's an opulent, full-bodied wine, showing its spice right up front. At the same time, it is well balanced, hinting at acidity at the end. Drink from 2015. \n","\n","\n","2 \n"," Tart and snappy, the flavors of lime flesh and rind dominate. Some green pineapple pokes through, with crisp acidity underscoring the flavors. The wine was all stainless-steel fermented. \n","\n","Packed with wood and fruit, this rich wine needs time. A juicy blend of 85% Merlot and 15% Cabernet Sauvignon, it's full of attractive acidity and blackberry fruits. The tannins do need to soften and round out, so drink from 2017. \n","\n","\n","Harvested at 22.6 degrees brix, this bright and spicy Dolcetto does not reach too far. It's tart, with fruit flavors that suggest wild berries and pomegranate, and tannins that are firm but never green. A fine quaffing wine. \n","\n","\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"qEC2Y-kWEB74","colab_type":"code","outputId":"4c9aeae1-a346-45df-89f9-c72b245632c3","executionInfo":{"status":"ok","timestamp":1576271593707,"user_tz":480,"elapsed":30447,"user":{"displayName":"Daniel McAndrew","photoUrl":"","userId":"01097539478049760558"}},"colab":{"base_uri":"https://localhost:8080/","height":665}},"source":["# Create red and white subsets and label which aggregated variety they belong to:\n","sample_df= df.sample(n=13000, random_state=42)\n","reds = [\"Noir\", \"Grenache\", \"Merlot\", \"Sangiovese\", \"Nebbiolo\", \"Tempranillo\", \"Cabernet Sauvignon\", \"Malbec\"]\n","whites = [\"Riesling\", \"Grigio\", \"Blanc\", \"Moscato\", \"Gewürztraminer\", \"Sémillon\", \"Viognier\", \"Chardonnay\"]\n","subsetRed = sample_df[sample_df['variety'].str.contains('|'.join(reds))]\n","subsetRed[\"variety\"] = \"Red\"\n","subsetWhite = sample_df[sample_df['variety'].str.contains('|'.join(whites))] \n","subsetWhite[\"variety\"] = \"White\"\n","#subsetRose = sample_df[sample_df['variety'].str.contains(\"Rosé\")] \n","#subsetRose[\"variety\"] = \"Rosé\"\n","\n","# t-SNE embed the subset:\n","subset = pd.concat([subsetRed, subsetWhite]) #,subsetRose])\n","data_embed=TSNE(n_components=2, perplexity=50, verbose=2, n_iter=500, method='barnes_hut').fit_transform(np.stack(subset[\"document_vector\"].to_numpy()))"],"execution_count":0,"outputs":[{"output_type":"stream","text":["/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:5: SettingWithCopyWarning: \n","A value is trying to be set on a copy of a slice from a DataFrame.\n","Try using .loc[row_indexer,col_indexer] = value instead\n","\n","See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n","  \"\"\"\n","/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:7: SettingWithCopyWarning: \n","A value is trying to be set on a copy of a slice from a DataFrame.\n","Try using .loc[row_indexer,col_indexer] = value instead\n","\n","See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n","  import sys\n"],"name":"stderr"},{"output_type":"stream","text":["[t-SNE] Computing 151 nearest neighbors...\n","[t-SNE] Indexed 6644 samples in 0.004s...\n","[t-SNE] Computed neighbors for 6644 samples in 0.792s...\n","[t-SNE] Computed conditional probabilities for sample 1000 / 6644\n","[t-SNE] Computed conditional probabilities for sample 2000 / 6644\n","[t-SNE] Computed conditional probabilities for sample 3000 / 6644\n","[t-SNE] Computed conditional probabilities for sample 4000 / 6644\n","[t-SNE] Computed conditional probabilities for sample 5000 / 6644\n","[t-SNE] Computed conditional probabilities for sample 6000 / 6644\n","[t-SNE] Computed conditional probabilities for sample 6644 / 6644\n","[t-SNE] Mean sigma: 0.051823\n","[t-SNE] Computed conditional probabilities in 0.750s\n","[t-SNE] Iteration 50: error = 85.4196396, gradient norm = 0.0127547 (50 iterations in 3.661s)\n","[t-SNE] Iteration 100: error = 78.3039551, gradient norm = 0.0002928 (50 iterations in 2.694s)\n","[t-SNE] Iteration 150: error = 78.2436523, gradient norm = 0.0000169 (50 iterations in 2.310s)\n","[t-SNE] Iteration 200: error = 78.2410965, gradient norm = 0.0000038 (50 iterations in 2.338s)\n","[t-SNE] Iteration 250: error = 78.2408981, gradient norm = 0.0000044 (50 iterations in 2.321s)\n","[t-SNE] KL divergence after 250 iterations with early exaggeration: 78.240898\n","[t-SNE] Iteration 300: error = 2.8333659, gradient norm = 0.0012407 (50 iterations in 2.592s)\n","[t-SNE] Iteration 350: error = 2.6270592, gradient norm = 0.0004884 (50 iterations in 3.181s)\n","[t-SNE] Iteration 400: error = 2.5195537, gradient norm = 0.0002698 (50 iterations in 3.136s)\n","[t-SNE] Iteration 450: error = 2.4527526, gradient norm = 0.0001881 (50 iterations in 3.113s)\n","[t-SNE] Iteration 500: error = 2.4080548, gradient norm = 0.0001312 (50 iterations in 3.101s)\n","[t-SNE] KL divergence after 500 iterations: 2.408055\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"2IrwoY7aFSIT","colab_type":"code","outputId":"bde0f867-ebdd-4781-c97e-f8ab51489106","executionInfo":{"status":"ok","timestamp":1576271596066,"user_tz":480,"elapsed":596,"user":{"displayName":"Daniel McAndrew","photoUrl":"","userId":"01097539478049760558"}},"colab":{"base_uri":"https://localhost:8080/","height":265}},"source":["## Vizualise\n","%matplotlib inline\n","from matplotlib import pyplot as plt\n","\n","x_axis= data_embed[:,0]\n","y_axis= data_embed[:,1]\n","\n","plt.scatter(x_axis, y_axis, s=2)\n","plt.show()"],"execution_count":0,"outputs":[{"output_type":"display_data","data":{"image/png":"iVBORw0KGgoAAAANSUhEUgAAAXoAAAD4CAYAAADiry33AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nBwZlXmA4AAAAAElFTkSuQmCC\n","text/plain":["<Figure size 432x288 with 1 Axes>"]},"metadata":{"tags":[]}}]},{"cell_type":"code","metadata":{"id":"eG0YbpchFuZb","colab_type":"code","colab":{}},"source":["# Create a tsv file for our embeded representation of our model:\n","d = {'x': data_embed[:,0], 'y': data_embed[:,1], 'variety' : subset[\"variety\"], \"points\" : subset[\"points\"], \"taster\" : subset[\"taster_name\"]}\n","df_embed = pd.DataFrame(data=d)\n","df_embed.head() \n","df_embed.to_csv('model2.tsv', sep='\\t', index=False)"],"execution_count":0,"outputs":[]},{"cell_type":"code","metadata":{"id":"xmdH-45VKWPK","colab_type":"code","colab":{}},"source":["#sample_df[\"variety\"].unique()\n","# df[\"taster_name\"].unique()"],"execution_count":0,"outputs":[]},{"cell_type":"code","metadata":{"id":"Iwb4ZP54LEbt","colab_type":"code","outputId":"d079d4a3-fb43-4a0e-8063-7877349557d2","executionInfo":{"status":"ok","timestamp":1576270305043,"user_tz":480,"elapsed":7283,"user":{"displayName":"Daniel McAndrew","photoUrl":"","userId":"01097539478049760558"}},"colab":{"base_uri":"https://localhost:8080/","height":737}},"source":["# Create subsets with just two varieties:\n","sample_df= df.sample(n=13000, random_state=42)\n","merls = [\"Merlot\"]\n","ries = [\"Riesling\"]\n","subsetmerls = sample_df[sample_df['variety'].str.contains('|'.join(merls))]\n","subsetmerls[\"variety\"] = \"Merlot\"\n","subsetries = sample_df[sample_df['variety'].str.contains('|'.join(ries))] \n","subsetries[\"variety\"] = \"Riesling\"\n","#subsetRose = sample_df[sample_df['variety'].str.contains(\"Rosé\")] \n","#subsetRose[\"variety\"] = \"Rosé\"\n","subset = pd.concat([subsetmerls, subsetries])#, subsetRose])\n","data_embed=TSNE(n_components=2, perplexity=50, verbose=2, method='barnes_hut').fit_transform(np.stack(subset[\"document_vector\"].to_numpy()))"],"execution_count":0,"outputs":[{"output_type":"stream","text":["/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:5: SettingWithCopyWarning: \n","A value is trying to be set on a copy of a slice from a DataFrame.\n","Try using .loc[row_indexer,col_indexer] = value instead\n","\n","See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n","  \"\"\"\n","/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:7: SettingWithCopyWarning: \n","A value is trying to be set on a copy of a slice from a DataFrame.\n","Try using .loc[row_indexer,col_indexer] = value instead\n","\n","See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n","  import sys\n"],"name":"stderr"},{"output_type":"stream","text":["[t-SNE] Computing 151 nearest neighbors...\n","[t-SNE] Indexed 876 samples in 0.001s...\n","[t-SNE] Computed neighbors for 876 samples in 0.037s...\n","[t-SNE] Computed conditional probabilities for sample 876 / 876\n","[t-SNE] Mean sigma: 0.072818\n","[t-SNE] Computed conditional probabilities in 0.094s\n","[t-SNE] Iteration 50: error = 60.7290001, gradient norm = 0.3244623 (50 iterations in 0.417s)\n","[t-SNE] Iteration 100: error = 60.1383591, gradient norm = 0.3300854 (50 iterations in 0.429s)\n","[t-SNE] Iteration 150: error = 60.4958916, gradient norm = 0.3117556 (50 iterations in 0.405s)\n","[t-SNE] Iteration 200: error = 61.2467461, gradient norm = 0.3163965 (50 iterations in 0.387s)\n","[t-SNE] Iteration 250: error = 60.2348518, gradient norm = 0.3202519 (50 iterations in 0.435s)\n","[t-SNE] KL divergence after 250 iterations with early exaggeration: 60.234852\n","[t-SNE] Iteration 300: error = 1.1302289, gradient norm = 0.0011643 (50 iterations in 0.335s)\n","[t-SNE] Iteration 350: error = 1.0902035, gradient norm = 0.0002916 (50 iterations in 0.314s)\n","[t-SNE] Iteration 400: error = 1.0797199, gradient norm = 0.0001689 (50 iterations in 0.312s)\n","[t-SNE] Iteration 450: error = 1.0761099, gradient norm = 0.0001877 (50 iterations in 0.311s)\n","[t-SNE] Iteration 500: error = 1.0744497, gradient norm = 0.0000752 (50 iterations in 0.311s)\n","[t-SNE] Iteration 550: error = 1.0737416, gradient norm = 0.0000783 (50 iterations in 0.342s)\n","[t-SNE] Iteration 600: error = 1.0735350, gradient norm = 0.0000607 (50 iterations in 0.312s)\n","[t-SNE] Iteration 650: error = 1.0733142, gradient norm = 0.0000655 (50 iterations in 0.314s)\n","[t-SNE] Iteration 700: error = 1.0729331, gradient norm = 0.0000526 (50 iterations in 0.311s)\n","[t-SNE] Iteration 750: error = 1.0728966, gradient norm = 0.0000477 (50 iterations in 0.317s)\n","[t-SNE] Iteration 800: error = 1.0728327, gradient norm = 0.0000437 (50 iterations in 0.318s)\n","[t-SNE] Iteration 850: error = 1.0726625, gradient norm = 0.0000491 (50 iterations in 0.327s)\n","[t-SNE] Iteration 900: error = 1.0725856, gradient norm = 0.0000425 (50 iterations in 0.321s)\n","[t-SNE] Iteration 950: error = 1.0726247, gradient norm = 0.0000414 (50 iterations in 0.316s)\n","[t-SNE] Iteration 1000: error = 1.0724441, gradient norm = 0.0000438 (50 iterations in 0.321s)\n","[t-SNE] KL divergence after 1000 iterations: 1.072444\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"LBr6Gz_fCfa9","colab_type":"code","colab":{}},"source":["from sklearn.model_selection import train_test_split\n","from sklearn.ensemble import RandomForestRegressor"],"execution_count":0,"outputs":[]},{"cell_type":"code","metadata":{"id":"CVP1-IHJpXeW","colab_type":"code","outputId":"e93ae407-1a6c-456f-bc33-18a949039ef5","executionInfo":{"status":"ok","timestamp":1576270310200,"user_tz":480,"elapsed":409,"user":{"displayName":"Daniel McAndrew","photoUrl":"","userId":"01097539478049760558"}},"colab":{"base_uri":"https://localhost:8080/","height":265}},"source":["boxplot = df.boxplot(column=['points'])"],"execution_count":0,"outputs":[{"output_type":"display_data","data":{"image/png":"iVBORw0KGgoAAAANSUhEUgAAAYAAAAD4CAYAAADlwTGnAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nIEkdZQBIUkcZAJLUUQaAJHXU/wUMQZC5XZRtQwAAAABJRU5ErkJggg==\n","text/plain":["<Figure size 432x288 with 1 Axes>"]},"metadata":{"tags":[]}}]},{"cell_type":"code","metadata":{"id":"l9TCsmiotg0D","colab_type":"code","colab":{}},"source":["df['points/price'] = (df[\"points\"] / df[\"price\"])"],"execution_count":0,"outputs":[]},{"cell_type":"code","metadata":{"id":"Rbd8aLeUt21Z","colab_type":"code","outputId":"854f536f-d2f7-4e0d-a2e7-bb4e6566bce5","executionInfo":{"status":"ok","timestamp":1576270314582,"user_tz":480,"elapsed":344,"user":{"displayName":"Daniel McAndrew","photoUrl":"","userId":"01097539478049760558"}},"colab":{"base_uri":"https://localhost:8080/","height":632}},"source":["df.head()"],"execution_count":0,"outputs":[{"output_type":"execute_result","data":{"text/html":["<div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>country</th>\n","      <th>description</th>\n","      <th>designation</th>\n","      <th>points</th>\n","      <th>price</th>\n","      <th>province</th>\n","      <th>region_1</th>\n","      <th>region_2</th>\n","      <th>taster_name</th>\n","      <th>taster_twitter_handle</th>\n","      <th>title</th>\n","      <th>variety</th>\n","      <th>winery</th>\n","      <th>idx</th>\n","      <th>document_vector</th>\n","      <th>closest_description</th>\n","      <th>points/price</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>Italy</td>\n","      <td>Aromas include tropical fruit, broom, brimston...</td>\n","      <td>Vulkà Bianco</td>\n","      <td>87</td>\n","      <td>NaN</td>\n","      <td>Sicily &amp; Sardinia</td>\n","      <td>Etna</td>\n","      <td>NaN</td>\n","      <td>Kerin O’Keefe</td>\n","      <td>@kerinokeefe</td>\n","      <td>Nicosia 2013 Vulkà Bianco  (Etna)</td>\n","      <td>White Blend</td>\n","      <td>Nicosia</td>\n","      <td>0</td>\n","      <td>[-0.031218965, -0.099459924, 0.010340899, 0.02...</td>\n","      <td>[127679, 98981, 26901, 43448, 35250]</td>\n","      <td>NaN</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>Portugal</td>\n","      <td>This is ripe and fruity, a wine that is smooth...</td>\n","      <td>Avidagos</td>\n","      <td>87</td>\n","      <td>15.0</td>\n","      <td>Douro</td>\n","      <td>NaN</td>\n","      <td>NaN</td>\n","      <td>Roger Voss</td>\n","      <td>@vossroger</td>\n","      <td>Quinta dos Avidagos 2011 Avidagos Red (Douro)</td>\n","      <td>Portuguese Red</td>\n","      <td>Quinta dos Avidagos</td>\n","      <td>1</td>\n","      <td>[0.018105369, -0.15248235, 0.049270216, 0.2182...</td>\n","      <td>[60658, 8648, 129235, 96895, 120482]</td>\n","      <td>5.800000</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>US</td>\n","      <td>Tart and snappy, the flavors of lime flesh and...</td>\n","      <td>NaN</td>\n","      <td>87</td>\n","      <td>14.0</td>\n","      <td>Oregon</td>\n","      <td>Willamette Valley</td>\n","      <td>Willamette Valley</td>\n","      <td>Paul Gregutt</td>\n","      <td>@paulgwine</td>\n","      <td>Rainstorm 2013 Pinot Gris (Willamette Valley)</td>\n","      <td>Pinot Gris</td>\n","      <td>Rainstorm</td>\n","      <td>2</td>\n","      <td>[0.038880985, -0.13466986, -0.00963622, 0.1377...</td>\n","      <td>[78546, 5169, 104867, 43029, 27643]</td>\n","      <td>6.214286</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>US</td>\n","      <td>Pineapple rind, lemon pith and orange blossom ...</td>\n","      <td>Reserve Late Harvest</td>\n","      <td>87</td>\n","      <td>13.0</td>\n","      <td>Michigan</td>\n","      <td>Lake Michigan Shore</td>\n","      <td>NaN</td>\n","      <td>Alexander Peartree</td>\n","      <td>NaN</td>\n","      <td>St. Julian 2013 Reserve Late Harvest Riesling ...</td>\n","      <td>Riesling</td>\n","      <td>St. Julian</td>\n","      <td>3</td>\n","      <td>[0.099366255, -0.2311258, 0.075494245, 0.16117...</td>\n","      <td>[125606, 43082, 124168, 69935, 28219]</td>\n","      <td>6.692308</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>US</td>\n","      <td>Much like the regular bottling from 2012, this...</td>\n","      <td>Vintner's Reserve Wild Child Block</td>\n","      <td>87</td>\n","      <td>65.0</td>\n","      <td>Oregon</td>\n","      <td>Willamette Valley</td>\n","      <td>Willamette Valley</td>\n","      <td>Paul Gregutt</td>\n","      <td>@paulgwine</td>\n","      <td>Sweet Cheeks 2012 Vintner's Reserve Wild Child...</td>\n","      <td>Pinot Noir</td>\n","      <td>Sweet Cheeks</td>\n","      <td>4</td>\n","      <td>[0.08297854, -0.25608635, 0.07158137, 0.157123...</td>\n","      <td>[96662, 71491, 120607, 75631, 33731]</td>\n","      <td>1.338462</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>"],"text/plain":["    country  ... points/price\n","0     Italy  ...          NaN\n","1  Portugal  ...     5.800000\n","2        US  ...     6.214286\n","3        US  ...     6.692308\n","4        US  ...     1.338462\n","\n","[5 rows x 17 columns]"]},"metadata":{"tags":[]},"execution_count":35}]},{"cell_type":"code","metadata":{"id":"wYvDHdK1xjg4","colab_type":"code","outputId":"c0f1d645-88b3-4346-c00d-ea5d2f19841d","executionInfo":{"status":"ok","timestamp":1576270316305,"user_tz":480,"elapsed":518,"user":{"displayName":"Daniel McAndrew","photoUrl":"","userId":"01097539478049760558"}},"colab":{"base_uri":"https://localhost:8080/","height":265}},"source":["hist = df[\"points/price\"].hist(bins=25, color=\"darkmagenta\")"],"execution_count":0,"outputs":[{"output_type":"display_data","data":{"image/png":"iVBORw0KGgoAAAANSUhEUgAAAYMAAAD4CAYAAAAO9oqkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n","text/plain":["<Figure size 432x288 with 1 Axes>"]},"metadata":{"tags":[]}}]},{"cell_type":"code","metadata":{"id":"lkX-BG1gybMG","colab_type":"code","outputId":"6c6a0509-d946-4aa3-e923-19c836d10c7d","executionInfo":{"status":"ok","timestamp":1576270318607,"user_tz":480,"elapsed":515,"user":{"displayName":"Daniel McAndrew","photoUrl":"","userId":"01097539478049760558"}},"colab":{"base_uri":"https://localhost:8080/","height":265}},"source":["boxplot = df.boxplot(column=['points/price'])"],"execution_count":0,"outputs":[{"output_type":"display_data","data":{"image/png":"iVBORw0KGgoAAAANSUhEUgAAAXAAAAD4CAYAAAD1jb0+AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nVCgLXJIK9X/WTT3RbT1HtgAAAABJRU5ErkJggg==\n","text/plain":["<Figure size 432x288 with 1 Axes>"]},"metadata":{"tags":[]}}]},{"cell_type":"code","metadata":{"id":"uJJUhaeiy8Dr","colab_type":"code","colab":{}},"source":["df_price_exists = df.dropna(subset=['price'])\n","X = df_price_exists[\"price\"]\n","y = df_price_exists[\"points\"]"],"execution_count":0,"outputs":[]},{"cell_type":"code","metadata":{"id":"2RqzKI525ygX","colab_type":"code","colab":{}},"source":["X_train, X_valid, y_train, y_valid = train_test_split(X, y, test_size=0.2, random_state=42)"],"execution_count":0,"outputs":[]},{"cell_type":"code","metadata":{"id":"YMhJE2u45z98","colab_type":"code","outputId":"52389927-adf9-4c58-8472-e27d298f4c9d","executionInfo":{"status":"ok","timestamp":1576272664046,"user_tz":480,"elapsed":2201,"user":{"displayName":"Daniel McAndrew","photoUrl":"","userId":"01097539478049760558"}},"colab":{"base_uri":"https://localhost:8080/","height":143}},"source":["regr = RandomForestRegressor(max_depth=3, random_state=42, n_estimators=100)\n","regr.fit(X_train.to_numpy().reshape(-1, 1), np.stack(y_train.to_numpy()))  "],"execution_count":0,"outputs":[{"output_type":"execute_result","data":{"text/plain":["RandomForestRegressor(bootstrap=True, criterion='mse', max_depth=3,\n","                      max_features='auto', max_leaf_nodes=None,\n","                      min_impurity_decrease=0.0, min_impurity_split=None,\n","                      min_samples_leaf=1, min_samples_split=2,\n","                      min_weight_fraction_leaf=0.0, n_estimators=100,\n","                      n_jobs=None, oob_score=False, random_state=42, verbose=0,\n","                      warm_start=False)"]},"metadata":{"tags":[]},"execution_count":86}]},{"cell_type":"code","metadata":{"id":"603AJWkYHipf","colab_type":"code","outputId":"e8482f3d-fc28-4b35-ea81-df85e1485f53","executionInfo":{"status":"ok","timestamp":1576272664049,"user_tz":480,"elapsed":2196,"user":{"displayName":"Daniel McAndrew","photoUrl":"","userId":"01097539478049760558"}},"colab":{"base_uri":"https://localhost:8080/","height":35}},"source":["y_pred = regr.predict(X_valid.to_numpy().reshape(-1, 1))\n","mean_absolute_error(y_pred, np.stack(y_valid.astype(float).to_numpy()))"],"execution_count":0,"outputs":[{"output_type":"execute_result","data":{"text/plain":["1.9201507133534814"]},"metadata":{"tags":[]},"execution_count":87}]},{"cell_type":"code","metadata":{"id":"SKSOYP6Bosgx","colab_type":"code","colab":{}},"source":["X2 = df[\"document_vector\"]\n","y2 = df[\"points\"]"],"execution_count":0,"outputs":[]},{"cell_type":"code","metadata":{"id":"FcYCvqi0Mf6p","colab_type":"code","colab":{}},"source":["X2_train, X2_valid, y2_train, y2_valid = train_test_split(X2, y2, test_size=0.2, random_state=42)"],"execution_count":0,"outputs":[]},{"cell_type":"code","metadata":{"id":"P32WZrwSMjKI","colab_type":"code","outputId":"058c6ac7-042a-4161-9bbf-b7546f0623c3","executionInfo":{"status":"ok","timestamp":1576272697149,"user_tz":480,"elapsed":35278,"user":{"displayName":"Daniel McAndrew","photoUrl":"","userId":"01097539478049760558"}},"colab":{"base_uri":"https://localhost:8080/","height":143}},"source":["regr = RandomForestRegressor(max_depth=3, random_state=42, n_estimators=100)\n","regr.fit(np.stack(X2_train.to_numpy()), np.stack(y2_train.to_numpy()))  "],"execution_count":0,"outputs":[{"output_type":"execute_result","data":{"text/plain":["RandomForestRegressor(bootstrap=True, criterion='mse', max_depth=3,\n","                      max_features='auto', max_leaf_nodes=None,\n","                      min_impurity_decrease=0.0, min_impurity_split=None,\n","                      min_samples_leaf=1, min_samples_split=2,\n","                      min_weight_fraction_leaf=0.0, n_estimators=100,\n","                      n_jobs=None, oob_score=False, random_state=42, verbose=0,\n","                      warm_start=False)"]},"metadata":{"tags":[]},"execution_count":90}]},{"cell_type":"code","metadata":{"id":"1yP7C1QJMqRx","colab_type":"code","outputId":"6c5c3f18-18ca-4a90-c78c-199f8990d88a","executionInfo":{"status":"ok","timestamp":1576272697303,"user_tz":480,"elapsed":35424,"user":{"displayName":"Daniel McAndrew","photoUrl":"","userId":"01097539478049760558"}},"colab":{"base_uri":"https://localhost:8080/","height":35}},"source":["y2_pred = regr.predict(np.stack(X2_valid.to_numpy()))\n","mean_absolute_error(y2_pred, np.stack(y2_valid.astype(float).to_numpy()))"],"execution_count":0,"outputs":[{"output_type":"execute_result","data":{"text/plain":["2.3765622707863803"]},"metadata":{"tags":[]},"execution_count":91}]},{"cell_type":"code","metadata":{"id":"z-Rl9l8eN_zl","colab_type":"code","colab":{}},"source":["dvec = df_price_exists[\"document_vector\"].tolist()"],"execution_count":0,"outputs":[]},{"cell_type":"code","metadata":{"id":"7SlH0XZhCjaz","colab_type":"code","colab":{}},"source":["X3 = []\n","for i in range(len(dvec)):\n","  X3.append(np.hstack((dvec[i], [df_price_exists[\"price\"].iloc[i]])))"],"execution_count":0,"outputs":[]},{"cell_type":"code","metadata":{"id":"VvN4wZamCoTZ","colab_type":"code","colab":{}},"source":["X3 = np.array(X3)"],"execution_count":0,"outputs":[]},{"cell_type":"code","metadata":{"id":"yM6f6-o8DlyL","colab_type":"code","colab":{}},"source":["X3_train, X3_valid, y3_train, y3_valid = train_test_split(X3, y, test_size=0.2, random_state=42)"],"execution_count":0,"outputs":[]},{"cell_type":"code","metadata":{"id":"V-AZzQSwDv1W","colab_type":"code","outputId":"02c2e5d3-098b-4651-c1cb-2af9ba4314d7","executionInfo":{"status":"ok","timestamp":1576272730424,"user_tz":480,"elapsed":68511,"user":{"displayName":"Daniel McAndrew","photoUrl":"","userId":"01097539478049760558"}},"colab":{"base_uri":"https://localhost:8080/","height":143}},"source":["regr = RandomForestRegressor(max_depth=3, random_state=42, n_estimators=100)\n","regr.fit(X3_train, np.stack(y_train.to_numpy()))  "],"execution_count":0,"outputs":[{"output_type":"execute_result","data":{"text/plain":["RandomForestRegressor(bootstrap=True, criterion='mse', max_depth=3,\n","                      max_features='auto', max_leaf_nodes=None,\n","                      min_impurity_decrease=0.0, min_impurity_split=None,\n","                      min_samples_leaf=1, min_samples_split=2,\n","                      min_weight_fraction_leaf=0.0, n_estimators=100,\n","                      n_jobs=None, oob_score=False, random_state=42, verbose=0,\n","                      warm_start=False)"]},"metadata":{"tags":[]},"execution_count":96}]},{"cell_type":"code","metadata":{"id":"rbaPZf8pEC6N","colab_type":"code","outputId":"20d5ca2f-ca2d-47c3-877f-76986930304b","executionInfo":{"status":"ok","timestamp":1576281427800,"user_tz":480,"elapsed":550,"user":{"displayName":"Daniel McAndrew","photoUrl":"","userId":"01097539478049760558"}},"colab":{"base_uri":"https://localhost:8080/","height":35}},"source":["y3_pred = regr.predict(np.stack(X3_valid))\n","mean_absolute_error(y3_pred, np.stack(y3_valid.astype(float).to_numpy()))"],"execution_count":0,"outputs":[{"output_type":"execute_result","data":{"text/plain":["1.9155034962787498"]},"metadata":{"tags":[]},"execution_count":99}]},{"cell_type":"code","metadata":{"id":"Ht8r3JZHEUVq","colab_type":"code","outputId":"2efc16e1-5b3f-447a-abd2-06f24f894a88","executionInfo":{"status":"ok","timestamp":1576272730641,"user_tz":480,"elapsed":68711,"user":{"displayName":"Daniel McAndrew","photoUrl":"","userId":"01097539478049760558"}},"colab":{"base_uri":"https://localhost:8080/","height":179}},"source":["df[\"points\"].describe()"],"execution_count":0,"outputs":[{"output_type":"execute_result","data":{"text/plain":["count    129971.000000\n","mean         88.447138\n","std           3.039730\n","min          80.000000\n","25%          86.000000\n","50%          88.000000\n","75%          91.000000\n","max         100.000000\n","Name: points, dtype: float64"]},"metadata":{"tags":[]},"execution_count":98}]}]}