added support to save viz

README.md

@@ -52,6 +52,9 @@ displayHTML(vis_html)
 
 ### Jupyter
 
+To save the visualization as html, provide the export file path: `save_path='./export.html'` for each visualizer.
+
+
 #### Dependency Parser
 ```python
 from sparknlp_display import DependencyParserVisualizer
@@ -64,7 +67,8 @@ pipeline_result = dp_pipeline.fullAnnotate(text)
 dependency_vis.display(pipeline_result[0], #should be the results of a single example, not the complete dataframe.
                        pos_col = 'pos', #specify the pos column
                        dependency_col = 'dependency', #specify the dependency column
-                       dependency_type_col = 'dependency_type' #specify the dependency type column
+                       dependency_type_col = 'dependency_type', #specify the dependency type column
+                       save_path='./export.html' # optional - to save viz as html. (default: None)
                        )
 ```
 
@@ -82,8 +86,9 @@ pipeline_result = ner_light_pipeline.fullAnnotate(text)
 
 ner_vis.display(pipeline_result[0], #should be the results of a single example, not the complete dataframe
                     label_col='entities', #specify the entity column
-                    document_col='document' #specify the document column (default: 'document')
-                    labels=['PER'] #only allow these labels to be displayed. (default: [] - all labels will be displayed)
+                    document_col='document', #specify the document column (default: 'document')
+                    labels=['PER'], #only allow these labels to be displayed. (default: [] - all labels will be displayed)
+                    save_path='./export.html' # optional - to save viz as html. (default: None)
                     )
 
 ## To set custom label colors:
@@ -104,8 +109,9 @@ pipeline_result = er_light_pipeline.fullAnnotate(text)
 
 er_vis.display(pipeline_result[0], #should be the results of a single example, not the complete dataframe
                label_col='entities', #specify the ner result column
-               resolution_col = 'resolution'
-               document_col='document' #specify the document column (default: 'document')
+               resolution_col = 'resolution',
+               document_col='document', #specify the document column (default: 'document')
+               save_path='./export.html' # optional - to save viz as html. (default: None)
                )
 
 ## To set custom label colors:
@@ -126,7 +132,8 @@ pipeline_result = re_light_pipeline.fullAnnotate(text)
 re_vis.display(pipeline_result[0], #should be the results of a single example, not the complete dataframe
                relation_col = 'relations', #specify relations column
                document_col = 'document', #specify document column
-               show_relations=True #display relation names on arrows (default: True)
+               show_relations=True, #display relation names on arrows (default: True)
+               save_path='./export.html' # optional - to save viz as html. (default: None)
                )
 
 ```
@@ -143,8 +150,9 @@ pipeline_result = ner_assertion_light_pipeline.fullAnnotate(text)
 
 assertion_vis.display(pipeline_result[0], 
                       label_col = 'entities', #specify the ner result column
-                      assertion_col = 'assertion' #specify assertion column
-                      document_col = 'document' #specify the document column (default: 'document')
+                      assertion_col = 'assertion', #specify assertion column
+                      document_col = 'document', #specify the document column (default: 'document')
+                      save_path='./export.html' # optional - to save viz as html. (default: None)
                       )
 
 ## To set custom label colors:

build/lib/sparknlp_display/VERSION

@@ -1 +1 @@
-1.8
+1.9

build/lib/sparknlp_display/assertion.py

@@ -158,7 +158,7 @@ def __display_ner(self, result, label_col, resolution_col, document_col, origina
 
         return html_output
 
-    def display(self, result, label_col, assertion_col, document_col='document', raw_text=None, return_html=False):
+    def display(self, result, label_col, assertion_col, document_col='document', raw_text=None, return_html=False, save_path=None):
         """Displays Assertion visualization. 
 
         Inputs:
@@ -174,9 +174,14 @@ def display(self, result, label_col, assertion_col, document_col='document', raw
         #self.__verifyInput(result, label_col, document_col, raw_text)
 
         html_content = self.__display_ner(result, label_col, assertion_col, document_col, raw_text)
+        html_content_save = style_config.STYLE_CONFIG_ENTITIES+ " "+html_content
+
+        if save_path != None:
+            with open(save_path, 'w') as f_:
+                f_.write(html_content_save)
 
         if return_html:
-            return style_config.STYLE_CONFIG_ENTITIES+ " "+html_content
+            return html_content_save
         else:
-            return display(HTML(style_config.STYLE_CONFIG_ENTITIES+ " "+html_content))
+            return display(HTML(html_content_save))
 

build/lib/sparknlp_display/dependency_parser.py

@@ -219,7 +219,7 @@ def __generate_graph(self, result_df):
         return dwg.tostring()
 
 
-    def display(self, res, pos_col, dependency_col, dependency_type_col=None, return_html=False):
+    def display(self, res, pos_col, dependency_col, dependency_type_col=None, return_html=False, save_path=None):
         """Displays NER visualization. 
 
         Inputs:
@@ -254,6 +254,11 @@ def display(self, res, pos_col, dependency_col, dependency_type_col=None, return
             df['dependency_type'] = ''
 
         html_content = self.__generate_graph(df)
+
+        if save_path != None:
+            with open(save_path, 'w') as f_:
+                f_.write(html_content)
+
         if return_html:
             return html_content
         else:

build/lib/sparknlp_display/entity_resolution.py

@@ -158,7 +158,7 @@ def __display_ner(self, result, label_col, resolution_col, document_col, origina
 
         return html_output
 
-    def display(self, result, label_col, resolution_col, document_col='document', raw_text=None, return_html=False):
+    def display(self, result, label_col, resolution_col, document_col='document', raw_text=None, return_html=False, save_path=None):
         """Displays NER visualization. 
 
         Inputs:
@@ -175,7 +175,13 @@ def display(self, result, label_col, resolution_col, document_col='document', ra
 
         html_content = self.__display_ner(result, label_col, resolution_col, document_col, raw_text)
 
+        html_content_save = style_config.STYLE_CONFIG_ENTITIES+ " "+html_content
+
+        if save_path != None:
+            with open(save_path, 'w') as f_:
+                f_.write(html_content_save)
+
         if return_html:
-            return style_config.STYLE_CONFIG_ENTITIES+ " "+html_content
+            return 
         else:
-            return display(HTML(style_config.STYLE_CONFIG_ENTITIES+ " "+html_content))
+            return display(HTML(html_content_save))

build/lib/sparknlp_display/ner.py

@@ -138,7 +138,7 @@ def __display_ner(self, result, label_col, document_col, original_text, labels_l
 
         return html_output
 
-    def display(self, result, label_col, document_col='document', raw_text=None, labels=None, return_html=False):
+    def display(self, result, label_col, document_col='document', raw_text=None, labels=None, return_html=False, save_path=None):
         """Displays NER visualization. 
         Inputs:
         result -- A Dataframe or dictionary.
@@ -153,8 +153,15 @@ def display(self, result, label_col, document_col='document', raw_text=None, lab
 
         html_content = self.__display_ner(result, label_col, document_col, raw_text, labels)
 
+        html_content_save = style_config.STYLE_CONFIG_ENTITIES+ " "+html_content
+
+        if save_path != None:
+            with open(save_path, 'w') as f_:
+                f_.write(html_content_save)
+
         if return_html:
-            return style_config.STYLE_CONFIG_ENTITIES+ " "+html_content
+            return html_content_save
         else:
-            return display(HTML(style_config.STYLE_CONFIG_ENTITIES+ " "+html_content))
+            return display(HTML(html_content_save))
+
 

build/lib/sparknlp_display/relation_extraction.py

@@ -375,7 +375,7 @@ def __gen_graph(self, rdf, selected_text, exclude_relations, show_relations):
 
         return dwg.tostring()
 
-    def display(self, result, relation_col, document_col='document', exclude_relations=['O'], show_relations=True, return_html=False):
+    def display(self, result, relation_col, document_col='document', exclude_relations=['O'], show_relations=True, return_html=False, save_path=None):
         """Displays Relation Extraction visualization. 
         Inputs:
         result -- A Dataframe or dictionary.
@@ -392,6 +392,10 @@ def display(self, result, relation_col, document_col='document', exclude_relatio
 
         html_content = self.__gen_graph(res, original_text, exclude_relations, show_relations)
 
+        if save_path != None:
+            with open(save_path, 'w') as f_:
+                f_.write(html_content)
+
         if return_html:
             return html_content
         else: