improving agg step

ruggleslab · Sep 18, 2019 · f03a7d5 · f03a7d5
1 parent 690ab13
commit f03a7d5
Show file tree

Hide file tree

Showing 2 changed files with 5 additions and 15 deletions.
diff --git a/blacksheep/_outlierTable.py b/blacksheep/_outlierTable.py
@@ -79,14 +79,12 @@ def _convert_to_counts(
     outlier_cols = [x + col_seps + col_outlier_suffix for x in samples]
 
     if aggregate:
-        df[agg_col] = [ind.split(ind_sep)[0] for ind in df.index]
+        df.index = [ind.split(ind_sep)[0] for ind in df.index]
+        df_inv = df == 0
+
         output_df = pd.DataFrame()
-        output_df[not_outlier_cols] = df.groupby(by=agg_col)[samples].agg(
-            lambda x: pd.Series(x == 0).sum()
-        )
-        output_df[outlier_cols] = df.groupby(by=agg_col)[samples].agg(
-            lambda x: pd.Series(x == 1).sum()
-        )
+        output_df[not_outlier_cols] = df_inv.groupby(level=0)[samples].sum()
+        output_df[outlier_cols] = df.groupby(level=0)[samples].sum()
     elif not aggregate:
         output_df = pd.DataFrame(index=df.index)
         output_df[outlier_cols] = df[samples]

diff --git a/blksheep/meta.yaml b/blksheep/meta.yaml
@@ -18,20 +18,12 @@ build:
 
 requirements:
   host:
-    - matplotlib >=3.1.0
-    - numpy >=1.16.4
-    - pandas >=0.24.2
     - pip
     - python >=3.6
-    - scikit-learn >=0.21.2
-    - scipy >=1.2.1
-    - seaborn >=0.9.0
-    - statsmodels >=0.10.0
   run:
     - matplotlib >=3.1.0
     - numpy >=1.16.4
     - pandas >=0.24.2
-    - python
     - scikit-learn >=0.21.2
     - scipy >=1.2.1
     - seaborn >=0.9.0