Update app.py

app.py

@@ -51,7 +51,7 @@ def load_data(active_card):
     # Columns specific to cards
     card_specific_cols = {
         'card1': ['FyWeek', 'Fy', 'State','Store','Address','Zipcode','City','Itemtype', 'Chaincode', 'Containercode', 'SalesVolume', 'UnitPrice', 'Sales'],
-        'card2': ['FyWeek', 'Fy', 'State','Store','Address','Zipcode','City','Itemtype', 'Chaincode', 'Containercode', 'SalesVolume', 'UnitPrice', 'Sales'],
+        # 'card2': ['FyWeek', 'Fy', 'State','Store','Address','Zipcode','City','Itemtype', 'Chaincode', 'Containercode', 'SalesVolume', 'UnitPrice', 'Sales'],
         'card3': ['FyWeek', 'Fy', 'State','Store','Address','Zipcode','City','Itemtype', 'Chaincode', 'Containercode', 'SalesVolume', 'UnitPrice', 'Sales'] # Added for PE calculation card
     }
 
@@ -114,7 +114,7 @@ if 'selected_feature' not in st.session_state:
     st.session_state['selected_feature'] = 'Chaincode'  # Default to 'Chain Code'
 
 # Card selection buttons with logic to reset session state on switch
-col1, col2, col3 = st.columns(3)
+col1, col3 = st.columns(2)
 with col1:
     if st.button("Sales Volume Trend"):
         st.session_state['active_card'] = 'card1'
@@ -124,14 +124,14 @@ with col1:
         st.session_state['selected_itemtype'] = None
         st.session_state['selected_containercode'] = None
 
-with col2:
-    if st.button("Sales Volume vs Median Unit Price Trend"):
-        st.session_state['active_card'] = 'card2'
-        # Reset selections when switching cards
-        st.session_state['selected_state'] = None
-        st.session_state['selected_chaincode'] = None
-        st.session_state['selected_itemtype'] = None
-        st.session_state['selected_containercode'] = None
+# with col2:
+#     if st.button("Sales Volume vs Median Unit Price Trend"):
+#         st.session_state['active_card'] = 'card2'
+#         # Reset selections when switching cards
+#         st.session_state['selected_state'] = None
+#         st.session_state['selected_chaincode'] = None
+#         st.session_state['selected_itemtype'] = None
+#         st.session_state['selected_containercode'] = None
 
 with col3:
     if st.button("Price Elasticity Coefficient Trend YoY"):
@@ -327,110 +327,110 @@ if st.session_state['active_card'] == 'card1':
 
 
 ########################################### CARD #2 ####################################################
-if st.session_state['active_card'] == 'card2':
-    # Identify the top 10 Itemtypes based on total SalesVolume
-    top_10_itemtypes = df.groupby('Itemtype')['SalesVolume'].sum().nlargest(10).index
+# if st.session_state['active_card'] == 'card2':
+#     # Identify the top 10 Itemtypes based on total SalesVolume
+#     top_10_itemtypes = df.groupby('Itemtype')['SalesVolume'].sum().nlargest(10).index
 
-    # Filter the DataFrame to include only the top 10 Itemtypes
-    df = df[df['Itemtype'].isin(top_10_itemtypes)]
-    # Dropdown to select item type (using session_state)
-    st.session_state['selected_item_type'] = st.selectbox(
-        'Select Item Type', df['Itemtype'].unique(),
-        index=list(df['Itemtype'].unique()).index(st.session_state['selected_item_type']))
+#     # Filter the DataFrame to include only the top 10 Itemtypes
+#     df = df[df['Itemtype'].isin(top_10_itemtypes)]
+#     # Dropdown to select item type (using session_state)
+#     st.session_state['selected_item_type'] = st.selectbox(
+#         'Select Item Type', df['Itemtype'].unique(),
+#         index=list(df['Itemtype'].unique()).index(st.session_state['selected_item_type']))
 
-    # Dropdown to select the grouping category (container code, chain code, or state)
-    group_by_option = st.selectbox('Group by', ['Containercode', 'Chaincode', 'State','Region'])
+#     # Dropdown to select the grouping category (container code, chain code, or state)
+#     group_by_option = st.selectbox('Group by', ['Containercode', 'Chaincode', 'State','Region'])
 
-    # Multi-select checkbox to select multiple years
-    selected_years = st.multiselect('Select Year(s)', [2021, 2022, 2023, 2024], default=[2021])
+#     # Multi-select checkbox to select multiple years
+#     selected_years = st.multiselect('Select Year(s)', [2021, 2022, 2023, 2024], default=[2021])
 
-    st.subheader(f"Sales Volume & Unit Price Correlation for {group_by_option} in {', '.join(map(str, selected_years))}")
+#     st.subheader(f"Sales Volume & Unit Price Correlation for {group_by_option} in {', '.join(map(str, selected_years))}")
 
-    # Convert 'Dt' column to datetime
-    df['Dt'] = pd.to_datetime(df['Dt'], errors='coerce')
-    df['Promo'] = np.where(df['Dt'].dt.month.astype(str).isin(['3', '4', '5', '6']), 'Promo', 'NoPromo')
-    df["Promo"] = df["Promo"].astype("category")
+#     # Convert 'Dt' column to datetime
+#     df['Dt'] = pd.to_datetime(df['Dt'], errors='coerce')
+#     df['Promo'] = np.where(df['Dt'].dt.month.astype(str).isin(['3', '4', '5', '6']), 'Promo', 'NoPromo')
+#     df["Promo"] = df["Promo"].astype("category")
 
-    # Filter the dataframe based on the selected item type and selected years
-    filtered_df = df[(df['Itemtype'] == st.session_state['selected_item_type']) & (df['Dt'].dt.year.isin(selected_years))]
+#     # Filter the dataframe based on the selected item type and selected years
+#     filtered_df = df[(df['Itemtype'] == st.session_state['selected_item_type']) & (df['Dt'].dt.year.isin(selected_years))]
 
-    # Find the top 3 values based on total SalesVolume in the selected grouping category
-    top_3_values = filtered_df.groupby(group_by_option, observed=True)['SalesVolume'].sum().nlargest(3).index
+#     # Find the top 3 values based on total SalesVolume in the selected grouping category
+#     top_3_values = filtered_df.groupby(group_by_option, observed=True)['SalesVolume'].sum().nlargest(3).index
 
-    # Filter the data for only the top 3 values
-    top_group_data = filtered_df[filtered_df[group_by_option].isin(top_3_values)]
+#     # Filter the data for only the top 3 values
+#     top_group_data = filtered_df[filtered_df[group_by_option].isin(top_3_values)]
     
-    # Aggregate data
-    agg_df = top_group_data.groupby([group_by_option, 'Year', 'Week', 'Dt'], observed=True).agg({
-        'SalesVolume': 'sum',
-        'UnitPrice': 'mean'
-    }).reset_index()
-
-    # Create a new column 'week-year' for X-axis labels
-    agg_df['week-year'] = agg_df['Dt'].dt.strftime('%U-%Y')
-
-    # Loop through the top 3 values and create separate plots using Plotly
-    for value in top_3_values:
-        value_data = agg_df[agg_df[group_by_option] == value]
-        # Assuming you have 'value_data' from your previous code
-        mean_sales_volume = value_data['SalesVolume'].mean()
-        mean_unit_price = value_data['UnitPrice'].mean()
-
-        # Create a Plotly figure
-        fig = go.Figure()
-
-        # Add SalesVolume trace
-        fig.add_trace(go.Scatter(
-            x=value_data['week-year'],
-            y=value_data['SalesVolume'],
-            mode='lines+markers',
-            name='SalesVolume',
-            line=dict(color='blue'),
-            hovertemplate='SalesVolume: %{y}<br>Week-Year: %{x}'
-        ))
-
-        # Add UnitPrice trace on a secondary Y-axis
-        fig.add_trace(go.Scatter(
-            x=value_data['week-year'],
-            y=value_data['UnitPrice'],
-            mode='lines+markers',
-            name='UnitPrice',
-            line=dict(color='green'),
-            yaxis='y2',
-            hovertemplate='UnitPrice: %{y}<br>Week-Year: %{x}'
-        ))
-        # Add mean line for SalesVolume
-        fig.add_shape(type="line",
-                    x0=value_data['week-year'].min(), x1=value_data['week-year'].max(),
-                    y0=mean_sales_volume, y1=mean_sales_volume,
-                    line=dict(color="blue", width=2, dash="dash"),
-                    xref='x', yref='y')
-
-        # Add mean line for UnitPrice (on secondary Y-axis)
-        fig.add_shape(type="line",
-                    x0=value_data['week-year'].min(), x1=value_data['week-year'].max(),
-                    y0=mean_unit_price, y1=mean_unit_price,
-                    line=dict(color="green", width=2, dash="dash"),
-                    xref='x', yref='y2')
-
-        # Update layout for dual axes
-        fig.update_layout(
-            template='plotly_white',
-            title=f"SalesVolume and UnitPrice - {value} ({group_by_option})",
-            xaxis_title='Week-Year',
-            yaxis_title='Sales Volume',
-            yaxis2=dict(title='UnitPrice', overlaying='y', side='right'),
-            legend=dict(x=0.9, y=1.15),
-            hovermode="x unified",  # Show both values in a tooltip
-            height=600,
-            margin=dict(l=50, r=50, t=50, b=50)
-        )
-
-        # Rotate X-axis labels
-        fig.update_xaxes(tickangle=90)
-
-        # Display the Plotly figure in Streamlit
-        st.plotly_chart(fig, use_container_width=True)
+#     # Aggregate data
+#     agg_df = top_group_data.groupby([group_by_option, 'Year', 'Week', 'Dt'], observed=True).agg({
+#         'SalesVolume': 'sum',
+#         'UnitPrice': 'mean'
+#     }).reset_index()
+
+#     # Create a new column 'week-year' for X-axis labels
+#     agg_df['week-year'] = agg_df['Dt'].dt.strftime('%U-%Y')
+
+#     # Loop through the top 3 values and create separate plots using Plotly
+#     for value in top_3_values:
+#         value_data = agg_df[agg_df[group_by_option] == value]
+#         # Assuming you have 'value_data' from your previous code
+#         mean_sales_volume = value_data['SalesVolume'].mean()
+#         mean_unit_price = value_data['UnitPrice'].mean()
+
+#         # Create a Plotly figure
+#         fig = go.Figure()
+
+#         # Add SalesVolume trace
+#         fig.add_trace(go.Scatter(
+#             x=value_data['week-year'],
+#             y=value_data['SalesVolume'],
+#             mode='lines+markers',
+#             name='SalesVolume',
+#             line=dict(color='blue'),
+#             hovertemplate='SalesVolume: %{y}<br>Week-Year: %{x}'
+#         ))
+
+#         # Add UnitPrice trace on a secondary Y-axis
+#         fig.add_trace(go.Scatter(
+#             x=value_data['week-year'],
+#             y=value_data['UnitPrice'],
+#             mode='lines+markers',
+#             name='UnitPrice',
+#             line=dict(color='green'),
+#             yaxis='y2',
+#             hovertemplate='UnitPrice: %{y}<br>Week-Year: %{x}'
+#         ))
+#         # Add mean line for SalesVolume
+#         fig.add_shape(type="line",
+#                     x0=value_data['week-year'].min(), x1=value_data['week-year'].max(),
+#                     y0=mean_sales_volume, y1=mean_sales_volume,
+#                     line=dict(color="blue", width=2, dash="dash"),
+#                     xref='x', yref='y')
+
+#         # Add mean line for UnitPrice (on secondary Y-axis)
+#         fig.add_shape(type="line",
+#                     x0=value_data['week-year'].min(), x1=value_data['week-year'].max(),
+#                     y0=mean_unit_price, y1=mean_unit_price,
+#                     line=dict(color="green", width=2, dash="dash"),
+#                     xref='x', yref='y2')
+
+#         # Update layout for dual axes
+#         fig.update_layout(
+#             template='plotly_white',
+#             title=f"SalesVolume and UnitPrice - {value} ({group_by_option})",
+#             xaxis_title='Week-Year',
+#             yaxis_title='Sales Volume',
+#             yaxis2=dict(title='UnitPrice', overlaying='y', side='right'),
+#             legend=dict(x=0.9, y=1.15),
+#             hovermode="x unified",  # Show both values in a tooltip
+#             height=600,
+#             margin=dict(l=50, r=50, t=50, b=50)
+#         )
+
+#         # Rotate X-axis labels
+#         fig.update_xaxes(tickangle=90)
+
+#         # Display the Plotly figure in Streamlit
+#         st.plotly_chart(fig, use_container_width=True)
 
 ################################
 if st.session_state['active_card'] == 'card3':
@@ -466,6 +466,11 @@ if st.session_state['active_card'] == 'card3':
 
     # Drop rows where PE_Coeff is NaN (optional)
     agg_df_filtered = agg_df_filtered.dropna(subset=['PE_Coeff'])
+    agg_df_filtered = agg_df_filtered.rename(columns={
+    'SalesVolume_pct_change': 'SlVol%change',
+    'UnitPrice_pct_change': 'UnPr%change',
+    })
+    agg_df_filtered = agg_df_filtered.reset_index(drop=True)
     st.dataframe(agg_df_filtered)
     st.write(agg_df_filtered.shape)
     # Extract values for the current and previous years from row 1 and row 2 of the dataframe
@@ -485,25 +490,53 @@ if st.session_state['active_card'] == 'card3':
     # Calculate PE Coefficient
     pe_coeff = sales_volume_pct / unit_price_pct
 
-    # Render LaTeX formulas with dynamic values using st.latex
+    st.markdown(f'''### Calculations for Price Elasticity Coefficient''')
     st.latex(rf"""
-    \text{{Unit Price \% Change}} = \frac{{\text{{Unit Price in Current Year}} - \text{{Unit Price in Previous Year}}}}{{\text{{Unit Price in Previous Year}}}} \times 100 = \frac{{{unit_price_current_year:.2f} - {unit_price_previous_year:.2f}}}{{{unit_price_previous_year:.2f}}} \times 100 = {unit_price_pct:.2f}\%
+    \text{{Unit Price \% Change}} = \frac{{{unit_price_current_year:.2f} - {unit_price_previous_year:.2f}}}{{{unit_price_previous_year:.2f}}} \times 100 = {unit_price_pct:.2f}\%
     """)
 
+    # Sales Volume % Change
     st.latex(rf"""
-    \text{{Sales Volume \% Change}} = \frac{{\text{{Sales Volume in Current Year}} - \text{{Sales Volume in Previous Year}}}}{{\text{{Sales Volume in Previous Year}}}} \times 100 = \frac{{{sales_volume_current_year:.2f} - {sales_volume_previous_year:.2f}}}{{{sales_volume_previous_year:.2f}}} \times 100 = {sales_volume_pct:.2f}\%
+    \text{{Sales Volume \% Change}} = \frac{{{sales_volume_current_year:.2f} - {sales_volume_previous_year:.2f}}}{{{sales_volume_previous_year:.2f}}} \times 100 = {sales_volume_pct:.2f}\%
     """)
 
+    # PE Coefficient
     st.latex(rf"""
-    \text{{PE Coefficient}} = \frac{{\text{{Sales Volume \% Change}}}}{{\text{{Unit Price \% Change}}}} = \frac{{{sales_volume_pct:.2f}}}{{{unit_price_pct:.2f}}} = {pe_coeff:.2f}
+    \text{{PE Coefficient}} = \frac{{{sales_volume_pct:.2f}}}{{{unit_price_pct:.2f}}} = {pe_coeff:.2f}
     """)
-    # Explanation Text
+
+    # Explanation for PE Coefficient Conditions
     st.markdown(f"""
-    ### Explanation of Calculations (for {current_year_row['Region']}, {current_year_row['Itemtype']}, FY {current_year_row['Fy']}):
-    - **Unit Price Percentage Change**: The percentage change in unit price YOY. For FY {current_year_row['Fy']}, the unit price changed by **{unit_price_pct:.2f}%**.
-    - **Sales Volume Percentage Change**: The percentage change in sales volume YOY. For FY {current_year_row['Fy']}, the sales volume changed by **{sales_volume_pct:.2f}%**.
-    - **Price Elasticity (PE) Coefficient**: The PE coefficient for FY {current_year_row['Fy']} is **{pe_coeff:.2f}**.
+    ### Interpretation of Price Elasticity (PE) Coefficient:
+    The Price Elasticity (PE) coefficient reflects how sensitive sales volume is to changes in unit price.
+
+    - If the **PE coefficient is positive**:
+    1. When the price increases, sales volume increases.
+    2. When the price decreases, sales volume decreases.
+    
+    - If the **PE coefficient is negative**:
+    1. When the price increases, sales volume decreases.
+    2. When the price decreases, sales volume increases.
+    
     """)
+
+    # Dynamic analysis based on the calculated PE coefficient and signs of changes
+    if unit_price_pct > 0 and sales_volume_pct > 0:
+        st.warning(f"""
+        Both unit price and sales volume increased (refer first and second row of the table). The PE coefficient of **{pe_coeff:.2f}** indicates that for every 1% increase in unit price, sales volume increased by approximately **{pe_coeff:.2f}%**.
+        """)
+    elif unit_price_pct < 0 and sales_volume_pct < 0:
+        st.warning(f"""
+        Both unit price and sales volume decreased (refer first and second row of the table). The PE coefficient of **{pe_coeff:.2f}** suggests that for every 1% decrease in unit price, sales volume decreased by approximately **{pe_coeff:.2f}%**.
+        """)
+    elif unit_price_pct > 0 and sales_volume_pct < 0:
+        st.warning(f"""
+        The unit price increased while sales volume decreased (refer first and second row of the table). The negative PE coefficient of **{pe_coeff:.2f}** means that for every 1% increase in unit price, sales volume fell by approximately **{abs(pe_coeff):.2f}%**.
+        """)
+    elif unit_price_pct < 0 and sales_volume_pct > 0:
+        st.warning(f"""
+        The unit price decreased while sales volume increased (refer first and second row of the table). The negative PE coefficient of **{pe_coeff:.2f}** implies that for every 1% decrease in unit price, sales volume increased by approximately **{abs(pe_coeff):.2f}%**.
+        """)
     # Plot the PE Coefficient with Plotly
     fig = px.line(
         agg_df_filtered,