junos
/
rapids
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

Fix minor bugs of modeling.py: f1-macro and proba

pull/95/head
Meng Li 2020-08-16 16:08:51 -04:00
parent b9306612cb
commit 123b78d438
1 changed files with 3 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
src/models/modeling.py
View File

@ -117,10 +117,10 @@ for train_index, test_index in outer_cv.split(data_x):
# Inner cross validation
if min(targets_value_counts) >= 6:
# SMOTE requires n_neighbors <= n_samples, the default value of n_neighbors is 6
clf = GridSearchCV(estimator=createPipeline(model, "SMOTE"), param_grid=model_hyperparams, cv=inner_cv, scoring="f1_micro")
clf = GridSearchCV(estimator=createPipeline(model, "SMOTE"), param_grid=model_hyperparams, cv=inner_cv, scoring="f1_macro")
else:
# RandomOverSampler: over-sample the minority class(es) by picking samples at random with replacement.
clf = GridSearchCV(estimator=createPipeline(model, "RandomOverSampler"), param_grid=model_hyperparams, cv=inner_cv, scoring="f1_micro")
clf = GridSearchCV(estimator=createPipeline(model, "RandomOverSampler"), param_grid=model_hyperparams, cv=inner_cv, scoring="f1_macro")
clf.fit(train_x, train_y.values.ravel())
# Collect results and parameters
@ -129,10 +129,7 @@ for train_index, test_index in outer_cv.split(data_x):
pred_y = pred_y + cur_fold_pred
proba_of_two_categories = clf.predict_proba(test_x).tolist()
if cur_fold_pred[0]:
pred_y_prob = pred_y_prob + [row[proba_of_two_categories[0].index(max(proba_of_two_categories[0]))] for row in proba_of_two_categories]
else:
pred_y_prob = pred_y_prob + [row[proba_of_two_categories[0].index(min(proba_of_two_categories[0]))] for row in proba_of_two_categories]
pred_y_prob = pred_y_prob + [probabilities[clf.classes_.tolist().index(1)] for probabilities in proba_of_two_categories]
true_y = true_y + test_y.values.ravel().tolist()
pid = pid + test_y.index.tolist() # each test partition (fold) in the outer cv is a participant (LeaveOneOut cv)