change pipeline load dataframe due to api change

SeldonIO · Nov 23, 2016 · 49ec918 · 49ec918
1 parent 8ec1af9
commit 49ec918
Show file tree

Hide file tree

Showing 6 changed files with 11 additions and 5 deletions.
diff --git a/docker/examples/finefoods/scripts/create_model.py b/docker/examples/finefoods/scripts/create_model.py
@@ -38,7 +38,7 @@ def train(self,sample):
         p = Pipeline(transformers)
 
         pw = sutl.Pipeline_wrapper()
-        df = pw.create_dataframe([self.data_folder],df_format="csv")
+        df = pw.create_dataframe_from_files([self.data_folder],df_format="csv")
         if sample < 1.0:
             logger.info("sampling dataset to size %s ",sample)
             df = df.sample(frac=sample,random_state=1)

diff --git a/docker/examples/iris/keras/keras_pipeline.py b/docker/examples/iris/keras/keras_pipeline.py
@@ -30,7 +30,8 @@ def run_pipeline(events,models):
     p = Pipeline(transformers)
 
     pw = sutl.Pipeline_wrapper()
-    df = pw.create_dataframe(events)
+    print events
+    df = pw.create_dataframe_from_files(events)
     df2 = p.fit(df)
     pw.save_pipeline(p,models)
 

diff --git a/docker/examples/iris/vw/vw_pipeline.py b/docker/examples/iris/vw/vw_pipeline.py
@@ -14,7 +14,7 @@ def run_pipeline(events,models):
     p = Pipeline(transformers)
 
     pw = sutl.Pipeline_wrapper()
-    df = pw.create_dataframe(events)
+    df = pw.create_dataframe_from_files(events)
     df2 = p.fit(df)
     pw.save_pipeline(p,models)
 

diff --git a/docker/examples/iris/xgboost/xgb_pipeline.py b/docker/examples/iris/xgboost/xgb_pipeline.py
@@ -20,7 +20,7 @@ def run_pipeline(events,models):
     p = Pipeline(transformers)
 
     pw = sutl.Pipeline_wrapper()
-    df = pw.create_dataframe(events)
+    df = pw.create_dataframe_from_files(events)
     df2 = p.fit_transform(df)
     pw.save_pipeline(p,models)
     logger.info("cross validation scores %s",cv.get_scores())

diff --git a/docker/examples/iris/xgboost_rpc/xgb_pipeline.py b/docker/examples/iris/xgboost_rpc/xgb_pipeline.py
@@ -20,7 +20,7 @@ def run_pipeline(events,models):
     p = Pipeline(transformers)
 
     pw = sutl.Pipeline_wrapper()
-    df = pw.create_dataframe(events)
+    df = pw.create_dataframe_from_files(events)
     df2 = p.fit_transform(df)
     pw.save_pipeline(p,models)
     logger.info("cross validation scores %s",cv.get_scores())

diff --git a/python/seldon/pipeline/util.py b/python/seldon/pipeline/util.py
@@ -125,6 +125,11 @@ def _convert_dataframe(self,local_file,df_format,csv_dates=None,index_col=None):
             return pd.read_json(local_file,orient='records')
 
 
+    def create_dataframe_from_files(self,locations,df_format="json",csv_dates=None,index_col=None):
+        local_file= self.work_folder+"/data"
+        self._copy_features_locally(locations,local_file,df_format)
+        return self._convert_dataframe(local_file,df_format,csv_dates,index_col)
+
     def create_dataframe(self,data=None,df_format="json",csv_dates=None,index_col=None):
         """
         Create Pandas dataframe from external source