updated the req files. added more tests

atomsci/ddm/test/unit/test_compare_splits_plots.py

@@ -6,7 +6,7 @@
 import pandas as pd
 
 from matplotlib import pyplot as plt
-from atomsci.ddm.utils.compare_splits_plots import SplitStats
+from atomsci.ddm.utils.compare_splits_plots import SplitStats, split, parse_args
 from matplotcheck.base import PlotTester
 
 # --- Fixtures ---
@@ -157,4 +157,111 @@ def test_dist_hist_plot_train_v_valid(mock_data):
     assert ax is not None
     pt.assert_num_bins(1)   # Check if histogram has been plotted
     pt.assert_axis_label_contains("x", "Tanimoto distance")
-    pt.assert_axis_label_contains("y", "Proportion of compounds")
+    pt.assert_axis_label_contains("y", "Proportion of compounds")
+
+def test_split_function(mock_data):
+    """
+    Test the `split` function to ensure it correctly splits the dataset into training, test, and validation sets.
+    Args:
+        mock_data (tuple): A tuple containing the total dataframe and the split dataframe.
+    Asserts:
+        - The length of the training, test, and validation dataframes are as expected.
+        - The 'cmpd_id' values in each subset are correct.
+    """
+    total_df, split_df = mock_data
+    train_df, test_df, valid_df = split(total_df, split_df, id_col='cmpd_id')
+
+    assert len(train_df) == 2
+    assert len(test_df) == 2
+    assert len(valid_df) == 1
+
+    assert set(train_df['cmpd_id']) == {'cmpd1', 'cmpd3'}
+    assert set(test_df['cmpd_id']) == {'cmpd2', 'cmpd5'}
+    assert set(valid_df['cmpd_id']) == {'cmpd4'}
+
+def test_split_stats_initialization(mock_data):
+    """
+    Test the initialization of the `SplitStats` class to ensure it correctly processes the input data.
+    Args:
+        mock_data (tuple): A tuple containing the total dataframe and the split dataframe.
+    Asserts:
+        - The attributes of the `SplitStats` object are correctly initialized.
+    """
+    total_df, split_df = mock_data
+    ss = SplitStats(total_df, split_df, smiles_col='smiles', id_col='cmpd_id', response_cols=['response'])
+
+    assert ss.smiles_col == 'smiles'
+    assert ss.id_col == 'cmpd_id'
+    assert ss.response_cols == ['response']
+    assert ss.total_df.equals(total_df)
+    assert ss.split_df.equals(split_df)
+    assert len(ss.train_df) == 2
+    assert len(ss.test_df) == 2
+    assert len(ss.valid_df) == 1
+
+def test_print_stats(mock_data, capsys):
+    """
+    Test the `print_stats` method to ensure it correctly prints the statistics.
+    Args:
+        mock_data (tuple): A tuple containing the total dataframe and the split dataframe.
+        capsys: Pytest fixture to capture stdout and stderr.
+    Asserts:
+        - The printed output contains the expected statistics.
+    """
+    total_df, split_df = mock_data
+    ss = SplitStats(total_df, split_df, smiles_col='smiles', id_col='cmpd_id', response_cols=['response'])
+    ss.print_stats()
+
+    captured = capsys.readouterr()
+    assert "dist tvt mean" in captured.out
+    assert "dist tvv mean" in captured.out
+    assert "train frac mean" in captured.out
+    assert "test frac mean" in captured.out
+    assert "valid frac mean" in captured.out
+
+def test_split(mock_data):
+    """
+    Test the `split` function to ensure it correctly splits the dataset into training, test, and validation sets.
+        
+    Args:
+        mock_data (tuple): A tuple containing two DataFrames, `total_df` and `split_df`, which represent the complete dataset 
+                        and the split dataset respectively.
+    Asserts:
+        - The training set contains the correct compounds.
+        - The test set contains the correct compounds.
+        - The validation set contains the correct compounds.
+    """
+    total_df, split_df = mock_data
+    train_df, test_df, valid_df = split(total_df, split_df, id_col='cmpd_id')
+
+    # Check training set
+    assert set(train_df['cmpd_id']) == {'cmpd1', 'cmpd3'}
+    # Check test set
+    assert set(test_df['cmpd_id']) == {'cmpd2', 'cmpd5'}
+    # Check validation set
+    assert set(valid_df['cmpd_id']) == {'cmpd4'}
+
+    # Check that the splits are mutually exclusive and collectively exhaustive
+    all_ids = set(total_df['cmpd_id'])
+    split_ids = set(train_df['cmpd_id']).union(set(test_df['cmpd_id'])).union(set(valid_df['cmpd_id']))
+    assert all_ids == split_ids
+
+def test_parse_args(mocker):
+    """
+    Test the `parse_args` function to ensure it correctly parses command-line arguments.
+    Args:
+        mocker: Pytest mocker fixture to mock command-line arguments.
+    Asserts:
+        - The parsed arguments match the expected values.
+    """
+    mocker.patch('sys.argv', [
+        'compare_splits_plots.py', 'dataset.csv', 'id', 'smiles', 'split_a.csv', 'split_b.csv', 'output_dir'
+    ])
+    args = parse_args()
+
+    assert args.csv == 'dataset.csv'
+    assert args.id_col == 'id'
+    assert args.smiles_col == 'smiles'
+    assert args.split_a == 'split_a.csv'
+    assert args.split_b == 'split_b.csv'
+    assert args.output_dir == 'output_dir'

pip/cpu_requirements.txt

@@ -5,6 +5,8 @@
 # for LC developers: comment out pypi index url and use t$
 # --extra-index-url https://wci-repo.llnl.gov/repository/$
 
+-r dev_requirements.txt
+
 tensorflow-cpu~=2.14.0
 
 torch==2.0.1
@@ -38,10 +40,6 @@ maestrowf
 MolVS
 mordred
 
-pytest
-matplotcheck
-ipykernel
-
 deepchem==2.7.1
 rdkit==2024.3.5
 pyyaml==5.4.1

pip/cuda_requirements.txt

@@ -5,6 +5,8 @@
 #--extra-index-url https://wci-repo.llnl.gov/repository/pypi-group/simple
 -f https://data.dgl.ai/wheels/cu118/repo.html
 
+-r dev_requirements.txt
+
 tensorflow[and-cuda]~=2.14.0
 
 tensorrt
@@ -40,10 +42,6 @@ maestrowf
 MolVS
 mordred
 
-pytest
-matplotcheck
-ipykernel
-
 deepchem==2.7.1
 rdkit==2024.3.5
 

pip/dev_requirements.txt

@@ -5,8 +5,10 @@ jupyterlab
 maestrowf
 matplotlib
 matplotlib-venn
+notebook
 pytest
 pytest-cov
+pytest-mock
 pytest-xdist
 matplotcheck
 ruff

pip/docker_requirements.txt

@@ -5,6 +5,8 @@
 # for LC developers: comment out pypi index url and use t$
 # --extra-index-url https://wci-repo.llnl.gov/repository/$
 
+-r dev_requirements.txt
+
 tensorflow-cpu~=2.14.0
 
 torch==2.0.1
@@ -38,11 +40,6 @@ bravado
 maestrowf
 MolVS
 mordred
-pytest
-matplotcheck
-ipykernel
-jupyter
-jupyterlab
 
 deepchem==2.7.1
 rdkit==2024.3.5

pip/mchip_requirements.txt

@@ -5,6 +5,8 @@
 # for LC developers: comment out pypi index url and use t$
 # --extra-index-url https://wci-repo.llnl.gov/repository/$
 
+-r dev_requirements.txt
+
 # use tensorflow not tensorflow-cpu for arm. see https://www.tensorflow.org/install/pip#linux
 tensorflow~=2.14.0
 
@@ -39,12 +41,6 @@ maestrowf
 MolVS
 mordred
 
-pytest
-matplotcheck
-ipykernel
-notebook
-jupyterlab
-
 deepchem==2.7.1
 rdkit==2024.3.5
 pyyaml==5.4.1