Fixup the abundance filtering

gmrukwa · Dec 10, 2019 · aec664c · aec664c
1 parent 7a9405c
commit aec664c
Show file tree

Hide file tree

Showing 3 changed files with 10 additions and 7 deletions.
diff --git a/.github/workflows/deploy.yml b/.github/workflows/deploy.yml
@@ -12,7 +12,7 @@ on:
 env:
   MAJOR: ${{ 2 }}
   MINOR: ${{ 3 }}
-  FIXUP: ${{ 1 }}
+  FIXUP: ${{ 2 }}
   PACKAGE_INIT_FILE: ${{ 'divik/__init__.py' }}
   DOCKER_REPO: ${{ 'gmrukwa/divik' }}
   IS_ALPHA: ${{ github.event_name == 'pull_request' }}

diff --git a/divik/__init__.py b/divik/__init__.py
@@ -1,4 +1,4 @@
-__version__ = '2.3.1'
+__version__ = '2.3.2'
 
 from ._seeding import seeded
 from ._sklearn import DiviK

diff --git a/divik/_feature_selection/_specialized.py b/divik/_feature_selection/_specialized.py
@@ -2,7 +2,8 @@
 from sklearn.base import BaseEstimator
 from sklearn.feature_selection.base import SelectorMixin
 from ._gmm_selector import GMMSelector
-from ._outlier import OutlierOrTopSelector
+from ._outlier import OutlierSelector, OutlierOrTopSelector
+from ._percentage_selector import PercentageSelector
 
 
 class HighAbundanceAndVarianceSelector(BaseEstimator, SelectorMixin):
@@ -146,11 +147,13 @@ def fit(self, X, y=None):
         -------
         self
         """
-        self.abundance_selector_ = OutlierOrTopSelector(
+        self.abundance_selector_ = OutlierSelector(
             stat='mean', use_log=self.use_log,
-            keep_outliers=self.keep_outliers,
-            min_features_rate=self.min_features_rate,
-            p=1.0 - self.p).fit(X)
+            keep_outliers=False).fit(X)
+        if self.abundance_selector_.selected_.mean() < self.min_features_rate:
+            self.abundance_selector_ = PercentageSelector(
+                stat='mean', use_log=self.use_log, keep_top=True,
+                p=1.0 - self.p).fit(X)
         filtered = self.abundance_selector_.transform(X)
         self.selected_ = self.abundance_selector_.selected_.copy()