Update app.py

app.py

@@ -233,24 +233,23 @@ def create_embeddings(texts_to_embedd):
 
 def predict(text_input, sample_size_slider, reduce_sample_checkbox, progress=gr.Progress()):
 
-    
-    # get data.
+    print('getting data to project')
+    progress(0, desc="Starting...")
 
     query, params = openalex_url_to_pyalex_query(text_input)
     query_length = query.count()
+    print(f'Requesting {query_length} entries...')
     
     records = []
     total_pages = (query_length + 199) // 200  # Calculate total number of pages
-
+    per_page = 0.3 / total_pages
     
-    for i, record in progress.tqdm(enumerate(chain(*query.paginate(per_page=200)))):
+    for i, record in enumerate(chain(*query.paginate(per_page=200))):
         records.append(record)
         
-        # Calculate progress from 0 to 0.1
-        achieved_progress = min(0., (i + 1) / query_length * 0.1)
         
         # Update progress bar
-        progress(achieved_progress, desc="Getting queried data...")
+        progress(per_page * i, desc="Getting queried data...")
 
 
     
@@ -410,9 +409,11 @@ with gr.Blocks() as block:
     gr.Markdown("""
     ## Mapping OpenAlex-Queries
 
-    This is a tool to further interdisciplinary research – you are a neuroscientist who has used ..., What have the ... been doing with them. 
-    You're a philosopher of science who wonders where the concept of a fitness landscape has appeared...
+    Enter the URL to an OpenAlex-search below. It will take a few minutes, but then the result will be projected onto a map of the OA database as a whole.
     """)
+
+#    This is a tool to further interdisciplinary research – you are a neuroscientist who has used ..., What have the ... been doing with them. 
+#    You're a philosopher of science who wonders where the concept of a fitness landscape has appeared...
     
     with gr.Column():
         text_input = gr.Textbox(label="OpenAlex Fulltext-Search")
@@ -434,7 +435,8 @@ with gr.Blocks() as block:
 def setup_basemap_data():
     # get data.
     print("getting basemap data...")
-    basedata_df = load("100k_filtered_OA_sample_cluster_and_positions.bz")
+    #basedata_df = load("100k_filtered_OA_sample_cluster_and_positions.bz")
+    base_data_df =pickle.load(open('100k_filtered_OA_sample_cluster_and_positions.pkl', 'rb'))
     print(basedata_df)
     return basedata_df