bdext 0.1.61__tar.gz → 0.1.62__tar.gz

{bdext-0.1.61/bdext.egg-info → bdext-0.1.62}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: bdext
-Version: 0.1.61
+Version: 0.1.62
 Summary: Estimation of BDEISS-CT parameters from phylogenetic trees.
 Home-page: https://github.com/modpath/bdeissct
 Author: Anna Zhukova

{bdext-0.1.61 → bdext-0.1.62}/bdeissct_dl/bdeissct_model.py

@@ -129,5 +129,5 @@ for model in (BDEISSCT, BDEISSCT1, BDEISSCT2, BDEISSCT2000):
 
 
 
-CT_EPI_COLUMNS = [REPRODUCTIVE_NUMBER, INFECTION_DURATION, SAMPLING_FRACTION, UPSILON, X_C, F_E, F_S, X_S, LA]
+CT_EPI_COLUMNS = [REPRODUCTIVE_NUMBER, INFECTION_DURATION, SAMPLING_FRACTION, LA, F_E, F_S, X_S, UPSILON, X_C]
 CT_RATE_COLUMNS = [PSI, RHO]

{bdext-0.1.61 → bdext-0.1.62}/bdeissct_dl/estimator_ct.py

@@ -7,12 +7,7 @@ from bdeissct_dl.model_serializer import load_model_keras, load_scaler_numpy
 
 def predict_parameters(df, model_path=MODEL_PATH):
     feature_columns = CT_EPI_COLUMNS
-    x_indices = []
-    for i, col in enumerate(df.columns):
-        if col in feature_columns:
-            x_indices.append(i)
-
-    X = df.iloc[:, x_indices].to_numpy(dtype=float, na_value=0)
+    X = df.loc[:, feature_columns].to_numpy(dtype=float, na_value=0)
 
     # Standardization of the input features with a
     # standard scaler
@@ -58,7 +53,6 @@ def main():
                              )
     parser.add_argument('--log', default=None, type=str, help="output log file")
     parser.add_argument('--sumstats', default=None, type=str, help="input file(s) with epi parameters")
-    parser.add_argument('--ci', action='store_true', help="calculate CIs")
     params = parser.parse_args()
 
     df = pd.read_csv(params.sumstats)

{bdext-0.1.61 → bdext-0.1.62}/bdeissct_dl/scaler_fitting.py

@@ -10,16 +10,13 @@ from bdeissct_dl.model_serializer import save_scaler_joblib, save_scaler_numpy
 from bdeissct_dl.training import get_data_characteristics
 
 
-def fit_scalers(paths, x_indices, y_indices, scaler_x=None, scaler_y=None):
+def fit_scalers(paths, x_indices, scaler_x=None):
     # First pass: calculate mean and var
     for path in paths:
         df = pd.read_csv(path)
         if scaler_x:
             X = df.iloc[:, x_indices].to_numpy(dtype=float, na_value=0)
             scaler_x.partial_fit(X)
-        if scaler_y:
-            Y = df.iloc[:, y_indices].to_numpy(dtype=float, na_value=0)
-            scaler_y.partial_fit(Y)
 
 
 def main():
@@ -41,17 +38,14 @@ def main():
 
     os.makedirs(params.model_path, exist_ok=True)
 
-    scaler_x, scaler_y = StandardScaler(), None
-    x_indices, y_indices, _ = \
+    scaler_x = StandardScaler()
+    x_indices, _ = \
         get_data_characteristics(paths=params.train_data, target_columns=TARGET_COLUMNS_BDEISSCT)
-    fit_scalers(paths=params.train_data, x_indices=x_indices, y_indices=y_indices, scaler_x=scaler_x, scaler_y=scaler_y)
+    fit_scalers(paths=params.train_data, x_indices=x_indices, scaler_x=scaler_x)
 
     if scaler_x is not None:
         save_scaler_joblib(scaler_x, params.model_path, suffix='x')
         save_scaler_numpy(scaler_x, params.model_path, suffix='x')
-    if scaler_y is not None:
-        save_scaler_joblib(scaler_y, params.model_path, suffix='y')
-        save_scaler_numpy(scaler_y, params.model_path, suffix='y')
 
 
 if '__main__' == __name__:

{bdext-0.1.61 → bdext-0.1.62}/bdeissct_dl/train_ct.py

@@ -86,12 +86,12 @@ def main():
         np.random.shuffle(params.val_data)
 
 
-    x_indices, y_indices, y_col2index = get_data_characteristics(paths=params.train_data,
+    x_indices, y_col2index = get_data_characteristics(paths=params.train_data,
                                                                  feature_columns=feature_columns,
                                                                  target_columns=target_columns)
 
-    scaler_x, scaler_y = StandardScaler(), None
-    fit_scalers(paths=params.train_data, x_indices=x_indices, y_indices=y_indices, scaler_x=scaler_x, scaler_y=scaler_y)
+    scaler_x = StandardScaler()
+    fit_scalers(paths=params.train_data, x_indices=x_indices, scaler_x=scaler_x)
 
     if scaler_x is not None:
         save_scaler_joblib(scaler_x, params.model_path, suffix='ct.x')

{bdext-0.1.61 → bdext-0.1.62}/bdeissct_dl/training.py

@@ -54,20 +54,20 @@ def get_test_data(dfs=None, paths=None, scaler_x=None):
 
 
 def get_data_characteristics(paths, target_columns=TARGET_COLUMNS_BDCT, feature_columns=None):
-    x_indices = []
-    y_indices = []
-    col2index = {}
+    col2index_y = {}
+    col2index_x = {}
 
     df = pd.read_csv(paths[0])
-    feature_columns = set(get_X_columns(df.columns)) if feature_columns is None else set(feature_columns)
-    target_columns = set(target_columns) if target_columns is not None else set()
+    feature_columns = get_X_columns(df.columns) if feature_columns is None else feature_columns
+    feature_column_set = set(feature_columns)
+    target_columns = target_columns if target_columns is not None else []
+    target_column_set = set(target_columns)
     for i, col in enumerate(df.columns):
-        if col in feature_columns:
-            x_indices.append(i)
-        if col in target_columns:
-            y_indices.append(i)
-            col2index[col] = i
-    return x_indices, y_indices, col2index
+        if col in feature_column_set:
+            col2index_x[col] = i
+        if col in target_column_set:
+            col2index_y[col] = i
+    return [col2index_x[_] for _ in feature_columns], col2index_y
 
 
 def get_train_data(target_columns, columns_x, columns_y, file_pattern=None, filenames=None, scaler_x=None, \
@@ -192,7 +192,7 @@ def main():
         np.random.shuffle(params.val_data)
 
 
-    x_indices, y_indices, y_col2index = get_data_characteristics(paths=params.train_data, target_columns=target_columns)
+    x_indices, y_col2index = get_data_characteristics(paths=params.train_data, target_columns=target_columns)
 
     scaler_x = load_scaler_numpy(params.model_path, suffix='x')
 

{bdext-0.1.61 → bdext-0.1.62/bdext.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: bdext
-Version: 0.1.61
+Version: 0.1.62
 Summary: Estimation of BDEISS-CT parameters from phylogenetic trees.
 Home-page: https://github.com/modpath/bdeissct
 Author: Anna Zhukova

{bdext-0.1.61 → bdext-0.1.62}/bdext.egg-info/SOURCES.txt

@@ -2,7 +2,6 @@ LICENSE
 README.md
 setup.py
 bdeissct_dl/__init__.py
-bdeissct_dl/assess_R_CT.py
 bdeissct_dl/bdeissct_model.py
 bdeissct_dl/dl_model.py
 bdeissct_dl/estimator.py

{bdext-0.1.61 → bdext-0.1.62}/setup.py

@@ -9,12 +9,12 @@ setup(
     long_description_content_type='text/markdown',
     include_package_data=True,
     package_data={'bdeissct_dl': [os.path.join('..', 'README.md'),
-                            # os.path.join('models', '*', '*.keras'),
-                            # os.path.join('models', '*', '*.h5'),
-                            # os.path.join('models', '*', '*.json'),
-                            # os.path.join('models', '*', '*.txt'),
-                            # os.path.join('models', '*', '*.npy'),
-                            # os.path.join('models', '*', '*.gz'),
+                            os.path.join('models', '*', '*.keras'),
+                            os.path.join('models', '*', '*.h5'),
+                            os.path.join('models', '*', '*.json'),
+                            os.path.join('models', '*', '*.txt'),
+                            os.path.join('models', '*', '*.npy'),
+                            os.path.join('models', '*', '*.gz'),
                             os.path.join('..', 'LICENCE')]},
     classifiers=[
         'Development Status :: 4 - Beta',
@@ -24,7 +24,7 @@ setup(
         'Topic :: Software Development',
         'Topic :: Software Development :: Libraries :: Python Modules',
     ],
-    version='0.1.61',
+    version='0.1.62',
     description='Estimation of BDEISS-CT parameters from phylogenetic trees.',
     author='Anna Zhukova',
     author_email='anna.zhukova@pasteur.fr',

bdext-0.1.61/bdeissct_dl/assess_R_CT.py

@@ -1,21 +0,0 @@
-import numpy as np
-import pandas as pd
-
-from bdeissct_dl.bdeissct_model import REPRODUCTIVE_NUMBER, INFECTION_DURATION, RHO, F_E, F_S, X_S, UPSILON, X_C
-from bdeissct_dl.tree_encoder import SCALING_FACTOR
-
-for model in ('BD', 'BDCT', 'BDEI', 'BDEICT', 'BDSS', 'BDSSCT', 'BDEISS', 'BDEISSCT'):
-    df = pd.read_csv(f'/home/azhukova/projects/bdeissct_dl/simulations_bdeissct/test/200_500/{model}/trees.csv.xz')
-    print(model, REPRODUCTIVE_NUMBER, np.quantile(df[REPRODUCTIVE_NUMBER], [0, 0.5, 1]))
-    print(model, INFECTION_DURATION, np.quantile(df[INFECTION_DURATION] * df[SCALING_FACTOR], [0, 0.5, 1]))
-    print(model, RHO, np.quantile(df[RHO], [0, 0.5, 1]))
-    if 'EI' in model:
-        print(model, F_E, np.quantile(df[F_E], [0, 0.5, 1]))
-        print(df[df[F_E] > 1].index)
-    if 'SS' in model:
-        print(model, F_S, np.quantile(df[F_S], [0, 0.5, 1]))
-        print(model, X_S, np.quantile(df[X_S], [0, 0.5, 1]))
-    if 'CT' in model:
-        print(model, UPSILON, np.quantile(df[UPSILON], [0, 0.5, 1]))
-        print(model, X_C, np.quantile(df[X_C], [0, 0.5, 1]))
-    print('---')