aesthetics

src/streamlit_app.py

@@ -29,7 +29,6 @@ Justifying value to design change,0.15
 
 def main():
     # st.write('# Semantic Alignment App')
-    st.write('**Start**')
     tab_bp, tab_is, tab_matching = st.tabs(['Business Processes', 'Information Systems', 'BP-IS Matching'], default='BP-IS Matching')
 
     with tab_bp:
@@ -49,36 +48,31 @@ def main():
     with tab_matching:
         
         business_process_classes = bp_df['business_process_class'].unique()
-        
         information_system_classes = is_df['information_system_class'].unique()
-        st.write(business_process_classes)
+        # st.write(business_process_classes)
         # st.write(information_system_classes)
 
         embeddings_bp_classes = model.encode(business_process_classes)
         embeddings_is_classes = model.encode(information_system_classes)
-        st.write(embeddings_bp_classes.shape, embeddings_is_classes.shape)
+        # st.write(embeddings_bp_classes.shape, embeddings_is_classes.shape)
         similarities = model.similarity(embeddings_bp_classes, embeddings_is_classes)
         similarities_minmax = minmax.fit_transform(similarities)
-        st.write('similarities.shape', similarities_minmax.shape) # bp, is
-        st.write('similarities.max()', similarities_minmax.max()) # bp, is
-        st.write('similarities.min()', similarities_minmax.min()) # bp, is
+        # st.write('similarities.max()', similarities_minmax.max()) # bp, is
+        # st.write('similarities.min()', similarities_minmax.min()) # bp, is
 
         # matrix = np.random.randn(len(business_process_classes), len(information_system_classes))
         # matrix = minmax.fit_transform(matrix) * 100
         matrix_df = pd.DataFrame(similarities_minmax, columns=information_system_classes , index=business_process_classes)
         
-        st.write('### Alignment')
+        st.write('### Alignment Matrix')
         st.dataframe(matrix_df)
 
         # st.write('### Bigger than 70')
         # matrix_df[matrix_df>70]
 
-        st.write('### Max Score')
+        st.write('### Max Score for each BP')
         max_col = similarities_minmax.max(1)
         max_col_arg = similarities_minmax.argmax(1)
-        # max_col = similarities.max(1)
-        # st.write('max_col', max_col)
-
         # information_system_classes[max_col], business_process_classes
         s = pd.DataFrame({'information_system_classes':information_system_classes[max_col_arg], 'business_process_classes':business_process_classes, 'score':max_col })
         st.dataframe(s)
@@ -86,7 +80,7 @@ def main():
         st.write('### Pillars')
         pillar_weights = pd.read_csv(StringIO(pillar_weights_csv))
         # step 1
-        st.write("### Step 1")
+        st.write("### Step 1 (from ChatGPT)")
         st.dataframe(pillar_weights)
         
         unique_pillars = pillar_weights['Pillar']
@@ -102,11 +96,11 @@ def main():
         similarities_pillars_df.columns = is_bp_pairs
         
         # step 2
-        st.write("### Step 2")
+        st.write("### Step 2 - BP-IS pair and Pillars (Embedding Model)")
         st.write(similarities_pillars_df.T)
         
         # step 3
-        st.write("### Step 3")
+        st.write("### Step 3 - Normalized Strategic Score (*0<score<1*)")
         strategic_scores = {k:0 for k in similarities_pillars_df.columns}
         
         for i, row in similarities_pillars_df.T.iterrows():