Update pages/market_rent_estimation.py

pages/market_rent_estimation.py

@@ -11,207 +11,207 @@ import geopandas as gpd
 
 
 def main():
-    st.set_page_config(initial_sidebar_state="collapsed", layout="wide")
-    tab1, tab2 = st.tabs(["Comps", "Combined estimation"])#, "Market distributions", "Rexy"])
- 
-    distances_matrix = np.load('atlanta_matrix.npy')
-    df_properties = pd.read_csv("atlanta_data_new.csv", encoding='utf-8')
- 
-    # user's filter for property size is a range in percentage around picked property 
-    min_size = (100 + st.session_state['min_property_lease_size_perc']) / 100 * st.session_state['building_sf']
-    max_size = (100 + st.session_state['max_property_lease_size_perc']) / 100 * st.session_state['building_sf']
+    st.set_page_config(initial_sidebar_state="collapsed", layout="wide")
+    tab1, tab2 = st.tabs(["Comps", "Combined estimation"])#, "Market distributions", "Rexy"])
+
+    distances_matrix = np.load('atlanta_matrix.npy')
+    df_properties = pd.read_csv("atlanta_data_new.csv", encoding='utf-8')
+    
+    # user's filter for property size is a range in percentage around picked property 
+    min_size = (100 + st.session_state['min_property_lease_size_perc']) / 100 * st.session_state['building_sf']
+    max_size = (100 + st.session_state['max_property_lease_size_perc']) / 100 * st.session_state['building_sf']
 
     # user's filter for lease size is a range in percentage around picked lease 
-    lease_min_size = (100 + st.session_state['min_property_size_perc']) / 100 * st.session_state['rented_sf']
-    lease_max_size = (100 + st.session_state['max_property_size_perc']) / 100 * st.session_state['rented_sf']
+    lease_min_size = (100 + st.session_state['min_property_size_perc']) / 100 * st.session_state['rented_sf']
+    lease_max_size = (100 + st.session_state['max_property_size_perc']) / 100 * st.session_state['rented_sf']
 
     #apply all filters 
-    mask = (df_properties['rented_sf'] >= lease_min_size) & (df_properties['rented_sf'] <= lease_max_size) & (df_properties['building_sf'] >= min_size) & (df_properties['building_sf'] <= max_size) & (df_properties['months_since'] <= st.session_state['months_back']) 
-    df_properties_filtered = df_properties[mask]
-   
-    # create distances matrix for remaining rows
-    positions = np.nonzero(mask)[0]
-    distances_matrix_filtered = distances_matrix[np.ix_(positions, positions)]
- 
-    # row_index = df_properties_filtered.index.get_loc(st.session_state['user_select_value'])
-    df_properties_filtered = df_properties_filtered.reset_index()
-    row_index = (df_properties_filtered[df_properties_filtered['google_ola'] == st.session_state['user_select_value']].index)[0]
-    distances = distances_matrix_filtered[row_index]
-    sorted_indices = np.argsort(distances)
-    sorted_distances = distances[sorted_indices]
-    sorted_indexes = df_properties_filtered.index[sorted_indices]
-    reordered_df_properties_filtered = df_properties_filtered.loc[sorted_indexes]
- 
-       
-    #comps page
-    with tab1:
-        filtered_data = reordered_df_properties_filtered[["google_ola", "market_costar", "submarket_costar", "execution_date", "rented_sf", "building_sf", "year_built", "office_rate", "min_clear_height", "max_clear_height", "docks", "drive_ins", "rent_combined"]]#pd.concat([filtered_data2])
-        comps_scores = sorted_distances
- 
-        filtered_data.insert(loc=1, column='Similarity score', value=comps_scores)
-       
-        # Formatting the DataFrame
-        filtered_data['Similarity score'] = ((1 - filtered_data['Similarity score']) * 100).apply(lambda x: f"{x:.2f}")
-        filtered_data['execution_date'] = pd.to_datetime(filtered_data['execution_date']).dt.strftime('%m-%d-%Y')
-        filtered_data['LSF (sf)'] = filtered_data['rented_sf'].round(0).astype(int)
-        filtered_data['RSF (sf)'] = filtered_data['building_sf'].round(0).astype(int)
-        filtered_data['Year built'] = filtered_data['year_built'].astype(int)
-        filtered_data.loc[filtered_data['office_rate'].notna(), 'office_rate'] = (filtered_data.loc[filtered_data['office_rate'].notna(), 'office_rate'] * 100).apply(lambda x: f"{x:.2f} %")
- 
-        filtered_data.loc[filtered_data['min_clear_height'].notna(), 'min_clear_height'] = (filtered_data.loc[filtered_data['min_clear_height'].notna(), 'min_clear_height']).apply(lambda x: f"{int(x)}")
-        # filtered_data['Clear Height (feet)'] = filtered_data['min_clear_height'].round(0).astype(int)
+    mask = (df_properties['rented_sf'] >= lease_min_size) & (df_properties['rented_sf'] <= lease_max_size) & (df_properties['building_sf'] >= min_size) & (df_properties['building_sf'] <= max_size) & (df_properties['months_since'] <= st.session_state['months_back']) 
+    df_properties_filtered = df_properties[mask]
+    
+    # create distances matrix for remaining rows
+    positions = np.nonzero(mask)[0]
+    distances_matrix_filtered = distances_matrix[np.ix_(positions, positions)]
+    
+    # row_index = df_properties_filtered.index.get_loc(st.session_state['user_select_value'])
+    df_properties_filtered = df_properties_filtered.reset_index()
+    row_index = (df_properties_filtered[df_properties_filtered['google_ola'] == st.session_state['user_select_value']].index)[0]
+    distances = distances_matrix_filtered[row_index]
+    sorted_indices = np.argsort(distances)
+    sorted_distances = distances[sorted_indices]
+    sorted_indexes = df_properties_filtered.index[sorted_indices]
+    reordered_df_properties_filtered = df_properties_filtered.loc[sorted_indexes]
  
-        filtered_data.loc[filtered_data['docks'].notna(), 'docks'] = (filtered_data.loc[filtered_data['docks'].notna(), 'docks']).apply(lambda x: f"{int(x)}")
-        filtered_data.loc[filtered_data['drive_ins'].notna(), 'drive_ins'] = (filtered_data.loc[filtered_data['drive_ins'].notna(), 'drive_ins']).apply(lambda x: f"{int(x)}")
- 
-        # filtered_data['Docks (/10ksf)'] = filtered_data['docks'].astype(int)
-        # filtered_data['Doors (/10ksf)'] = filtered_data['drive_ins'].astype(int)
-        filtered_data['Rent (NNN)'] = filtered_data['rent_combined'].apply(lambda x: f"${x:.2f}")
        
-        # Dropping old columns and renaming headers
-        filtered_data = filtered_data.drop(columns=['rented_sf', 'building_sf', 'year_built', 'max_clear_height', 'rent_combined'])
-        filtered_data = filtered_data.rename(columns={
-            'google_ola': 'Address',
-            'office_rate': 'Office %',
-            'min_clear_height': 'Clear Height (feet)',
-            'docks': 'Docks (/10ksf)',
-            'drive_ins': 'Doors (/10ksf)',
-            'market_costar': 'Market',
-            'submarket_costar': 'Submarket'
-        })
-        filtered_data = filtered_data.sort_values(by="Similarity score", ascending=False)
+    #comps page
+    with tab1:
+        filtered_data = reordered_df_properties_filtered[["google_ola", "market_costar", "submarket_costar", "execution_date", "rented_sf", "building_sf", "year_built", "office_rate", "min_clear_height", "max_clear_height", "docks", "drive_ins", "rent_combined"]]#pd.concat([filtered_data2])
+        comps_scores = sorted_distances
+        
+        filtered_data.insert(loc=1, column='Similarity score', value=comps_scores)
+        
+        # Formatting the DataFrame
+        filtered_data['Similarity score'] = ((1 - filtered_data['Similarity score']) * 100).apply(lambda x: f"{x:.2f}")
+        filtered_data['execution_date'] = pd.to_datetime(filtered_data['execution_date']).dt.strftime('%m-%d-%Y')
+        filtered_data['LSF (sf)'] = filtered_data['rented_sf'].round(0).astype(int)
+        filtered_data['RSF (sf)'] = filtered_data['building_sf'].round(0).astype(int)
+        filtered_data['Year built'] = filtered_data['year_built'].astype(int)
+        filtered_data.loc[filtered_data['office_rate'].notna(), 'office_rate'] = (filtered_data.loc[filtered_data['office_rate'].notna(), 'office_rate'] * 100).apply(lambda x: f"{x:.2f} %")
+        
+        filtered_data.loc[filtered_data['min_clear_height'].notna(), 'min_clear_height'] = (filtered_data.loc[filtered_data['min_clear_height'].notna(), 'min_clear_height']).apply(lambda x: f"{int(x)}")
+        # filtered_data['Clear Height (feet)'] = filtered_data['min_clear_height'].round(0).astype(int)
+        
+        filtered_data.loc[filtered_data['docks'].notna(), 'docks'] = (filtered_data.loc[filtered_data['docks'].notna(), 'docks']).apply(lambda x: f"{int(x)}")
+        filtered_data.loc[filtered_data['drive_ins'].notna(), 'drive_ins'] = (filtered_data.loc[filtered_data['drive_ins'].notna(), 'drive_ins']).apply(lambda x: f"{int(x)}")
+        
+        # filtered_data['Docks (/10ksf)'] = filtered_data['docks'].astype(int)
+        # filtered_data['Doors (/10ksf)'] = filtered_data['drive_ins'].astype(int)
+        filtered_data['Rent (NNN)'] = filtered_data['rent_combined'].apply(lambda x: f"${x:.2f}")
+        
+        # Dropping old columns and renaming headers
+        filtered_data = filtered_data.drop(columns=['rented_sf', 'building_sf', 'year_built', 'max_clear_height', 'rent_combined'])
+        filtered_data = filtered_data.rename(columns={
+            'google_ola': 'Address',
+            'office_rate': 'Office %',
+            'min_clear_height': 'Clear Height (feet)',
+            'docks': 'Docks (/10ksf)',
+            'drive_ins': 'Doors (/10ksf)',
+            'market_costar': 'Market',
+            'submarket_costar': 'Submarket'
+        })
+        filtered_data = filtered_data.sort_values(by="Similarity score", ascending=False)
        
        
-        # Display the filtered data
-        col_1_1, col_1_2 = st.columns([2, 1])
-        with col_1_1:
-            filtered_data['Similarity score'] = pd.to_numeric(filtered_data['Similarity score'], errors='coerce')
-            filtered_data['LSF (sf)'] = pd.to_numeric(filtered_data['LSF (sf)'], errors='coerce')
-            st.write('Comps list:')
-            filtered_data_sorted = filtered_data.sort_values(by="Similarity score", ascending=False)
+        # Display the filtered data
+        col_1_1, col_1_2 = st.columns([2, 1])
+        with col_1_1:
+            filtered_data['Similarity score'] = pd.to_numeric(filtered_data['Similarity score'], errors='coerce')
+            filtered_data['LSF (sf)'] = pd.to_numeric(filtered_data['LSF (sf)'], errors='coerce')
+            st.write('Comps list:')
+            filtered_data_sorted = filtered_data.sort_values(by="Similarity score", ascending=False)
+            
+            # sort_column = st.selectbox('Select column to sort by:', filtered_data.columns[:])
+            # # display_df = st.table(filtered_data)
+            # if sort_column:
+            #     filtered_data_sorted = filtered_data.sort_values(by=sort_column, ascending=False)         
+            st.dataframe(filtered_data_sorted)
+ 
+        with col_1_2:
+        # Create a map object
+            m = folium.Map(width=500, height=440, location=(reordered_df_properties_filtered['lat'].mean(), reordered_df_properties_filtered['long'].mean()), zoom_start=9)
            
-            # sort_column = st.selectbox('Select column to sort by:', filtered_data.columns[:])
-            # # display_df = st.table(filtered_data)
-            # if sort_column:
-            #     filtered_data_sorted = filtered_data.sort_values(by=sort_column, ascending=False)         
-            st.dataframe(filtered_data_sorted)
- 
-        with col_1_2:
-        # Create a map object
-            m = folium.Map(width=500, height=440, location=(reordered_df_properties_filtered['lat'].mean(), reordered_df_properties_filtered['long'].mean()), zoom_start=9)
-           
-            # Add markers to the map
-            all_markers = folium.FeatureGroup(name='All Markers')
-            active_markers = folium.FeatureGroup(name='Active Markers', show=False)
-            inactive_markers = folium.FeatureGroup(name='Inactive Markers', show=False)
+            # Add markers to the map
+            all_markers = folium.FeatureGroup(name='All Markers')
+            active_markers = folium.FeatureGroup(name='Active Markers', show=False)
+            inactive_markers = folium.FeatureGroup(name='Inactive Markers', show=False)
                
-            for index, row in reordered_df_properties_filtered.iterrows():
-                status_color = 'green' if index==0 else 'red'
-                html_content = f"""
-                <div style="
-                    display: inline-block;
-                    background-color: white;
-                    border: 2px solid black;
-                    border-radius: 50%;
-                    width: 20px;
-                    height: 20px;
-                    text-align: center;
-                    line-height: 20px;
-                    font-size: 8pt;
-                    color: {status_color};
-                ">{index}</div>
-                """
+            for index, row in reordered_df_properties_filtered.iterrows():
+                status_color = 'green' if index==0 else 'red'
+                html_content = f"""
+                <div style="
+                    display: inline-block;
+                    background-color: white;
+                    border: 2px solid black;
+                    border-radius: 50%;
+                    width: 20px;
+                    height: 20px;
+                    text-align: center;
+                    line-height: 20px;
+                    font-size: 8pt;
+                    color: {status_color};
+                ">{index}</div>
+                """
                    
-                # Create a DivIcon with custom HTML content
-                icon = folium.DivIcon(html=html_content)
-                marker = folium.Marker([row['lat'], row['long']], popup=row['google_ola'], icon=icon).add_to(m)
+                # Create a DivIcon with custom HTML content
+                icon = folium.DivIcon(html=html_content)
+                marker = folium.Marker([row['lat'], row['long']], popup=row['google_ola'], icon=icon).add_to(m)
  
  
-            #add poligons on map
-            gdf = gpd.read_file('costar_sm_polygons.geojson')
-            gdf_Atlanta = gdf[gdf.full_submarket.str.contains("Atlanta")]
-            folium.GeoJson(data=gdf_Atlanta.geometry.to_json(), name='geojson').add_to(m)
+            #add poligons on map
+            gdf = gpd.read_file('costar_sm_polygons.geojson')
+            gdf_Atlanta = gdf[gdf.full_submarket.str.contains("Atlanta")]
+            folium.GeoJson(data=gdf_Atlanta.geometry.to_json(), name='geojson').add_to(m)
             for _, row in gdf_Atlanta.iterrows():
-                # Convert the row's geometry to GeoJSON
-                geojson = folium.GeoJson(data=row['geometry'].__geo_interface__,
-                                         style_function=lambda feature: {
-                                               'fillColor': '#00000000',  # Make the fill color transparent
-                                               'color': '#00000000',             # Set the border color
-                                               'weight': 2                  # Set the border width
-                                             },                                        
-                                        )
-                # Add a popup with the name
-                popup = folium.Popup(row['full_submarket'], parse_html=True)
+                # Convert the row's geometry to GeoJSON
+                geojson = folium.GeoJson(data=row['geometry'].__geo_interface__,
+                                            style_function=lambda feature: {
+                                            'fillColor': '#00000000',  # Make the fill color transparent
+                                            'color': '#00000000',             # Set the border color
+                                            'weight': 2                  # Set the border width
+                                            },                                        
+                                        )
+                # Add a popup with the name
+                popup = folium.Popup(row['full_submarket'], parse_html=True)
                
-                # Add the GeoJson and Popup to the map
-                geojson.add_child(popup).add_to(m)
+                # Add the GeoJson and Popup to the map
+                geojson.add_child(popup).add_to(m)
                    
-            # Add layer control to toggle marker visibility
-            folium.LayerControl().add_to(m)
+            # Add layer control to toggle marker visibility
+            folium.LayerControl().add_to(m)
            
-            # Render the map
-            folium_static(m)
+            # Render the map
+            folium_static(m)
  
-        back_to_serach_tab2 = st.button("Search page ")
-        if back_to_serach_tab2:
-            st.switch_page("app.py")
+        back_to_serach_tab2 = st.button("Search page ")
+        if back_to_serach_tab2:
+            st.switch_page("app.py")
            
-    with tab2:
-        st.title('Combined estimation')
-        if 'user_select_value' not in st.session_state:
-            st.session_state['user_select_value'] = ""
-        if 'year_buit' not in st.session_state:
-            st.session_state['year_buit'] = ""
-        if 'submarket_val' not in st.session_state:
-            st.session_state['submarket_val'] = ""
-        if 'market_val' not in st.session_state:
-            st.session_state['market_val'] = ""
- 
-        box_contents = [
-            {"header": "Address", "content": st.session_state['user_select_value']},
-            {"header": "LSF", "content": str(st.session_state['rented_sf'])},
-            {"header": "Sub-Market", "content": st.session_state['submarket_val']},
-            {"header": "Market", "content": st.session_state['market_val']}
-        ]
- 
-        # Divide the layout into four columns
-        col1, col2, col3, col4 = st.columns(4)
+    with tab2:
+        st.title('Combined estimation')
+        if 'user_select_value' not in st.session_state:
+            st.session_state['user_select_value'] = ""
+        if 'year_buit' not in st.session_state:
+            st.session_state['year_buit'] = ""
+        if 'submarket_val' not in st.session_state:
+            st.session_state['submarket_val'] = ""
+        if 'market_val' not in st.session_state:
+            st.session_state['market_val'] = ""
+ 
+        box_contents = [
+            {"header": "Address", "content": st.session_state['user_select_value']},
+            {"header": "LSF", "content": str(st.session_state['rented_sf'])},
+            {"header": "Sub-Market", "content": st.session_state['submarket_val']},
+            {"header": "Market", "content": st.session_state['market_val']}
+        ]
+ 
+        # Divide the layout into four columns
+        col1, col2, col3, col4 = st.columns(4)
        
-        for i, col in enumerate([col1, col2, col3, col4]):
-            col.markdown(f"""
-                <div style="padding: 20px; margin: 10px; text-align: center;">
-                    <h3 style="font-size: small;">{box_contents[i]['header']}</h3>
-                    <p style="font-size: small;">{box_contents[i]['content']}</p>
-                </div>
-            """, unsafe_allow_html=True)
+        for i, col in enumerate([col1, col2, col3, col4]):
+            col.markdown(f"""
+                    <div style="padding: 20px; margin: 10px; text-align: center;">
+                    <h3 style="font-size: small;">{box_contents[i]['header']}</h3>
+                    <p style="font-size: small;">{box_contents[i]['content']}</p>
+                    </div>
+            """, unsafe_allow_html=True)
            
-        # Add padding between the boxes and the slider
-        st.markdown('<style>div[data-testid="stBlock"]{margin-top: 20px;}</style>', unsafe_allow_html=True)
+        # Add padding between the boxes and the slider
+        st.markdown('<style>div[data-testid="stBlock"]{margin-top: 20px;}</style>', unsafe_allow_html=True)
        
-        def safe_eval(val):
-            try:
-                return ast.literal_eval(val)
-            except (ValueError, SyntaxError):
-                return val
+        # def safe_eval(val):
+        #     try:
+        #         return ast.literal_eval(val)
+        #     except (ValueError, SyntaxError):
+        #         return val
                 
-        average_rent = reordered_df_properties_filtered['rent_combined'].mean()
+        average_rent = reordered_df_properties_filtered['rent_combined'].mean()
        
-        x1 ="Comps " + str(average_rent)
-        x3="Rent " + str(st.session_state['prediction'])
+        x1 ="Comps " + str(average_rent)
+        x3="Rent " + str(st.session_state['prediction'])
        
-        # x2 = (st.session_state['prediction']+average_rent)/2
+        # x2 = (st.session_state['prediction']+average_rent)/2
        
-        pick = st.select_slider(
-        "Combined estimation ",
-        options=[x1, x3],
-        value=x1)
+        pick = st.select_slider(
+        "Combined estimation ",
+        options=[x1, x3],
+        value=x1)
        
-        back_to_serach_tab1 = st.button("Search page")
-        if back_to_serach_tab1:
-            st.switch_page("app.py")
+        back_to_serach_tab1 = st.button("Search page")
+        if back_to_serach_tab1:
+            st.switch_page("app.py")
  
            
  
 if __name__ == "__main__":
-    main()
+    main()