fixed float to int

Author: jjacobson95

build/beatAML/GetBeatAML.py

@@ -466,6 +466,11 @@ def map_and_combine(df, data_type, entrez_map_file, improve_map_file, map_file=N
                          right_on='other_id',
                          how='left')
     mapped_df.insert(0, 'improve_sample_id', mapped_df.pop('improve_sample_id'))
+    
+    print(mapped_df.to_string())
+    mapped_df['improve_sample_id'] = mapped_df['improve_sample_id'].astype(int)
+    mapped_df['entrez_id'] = mapped_df['entrez_id'].fillna(0)
+    mapped_df['entrez_id'] = mapped_df['entrez_id'].astype(int)
     mapped_df['source'] = 'synapse'
     mapped_df['study'] = 'BeatAML'
 

build/cptac/getCptacData.py

@@ -380,11 +380,15 @@ def main():
                     dat_files[dtype_key] = fdf2
                 else:
                     dat_files[dtype_key] = fdf.dropna()
+                
                 print(dtype_key)
 
         # Now concatenate all the cancers into a single file
         for dtype_key, df in dat_files.items():
             print('Saving ' + "cptac_" + dtype_key + '.csv.gz' + ' file')
+            print(df.to_string())
+            df['entrez_id'] = df['entrez_id'].fillna(0)
+            df['entrez_id'] = df['entrez_id'].astype(int)
             df.to_csv("/tmp/" + "cptac_" + dtype_key + '.csv.gz', sep=',', index=False, compression='gzip')
 
 if __name__ == '__main__':