Update app.py

app.py

@@ -492,6 +492,24 @@ def process_dataframe(df):
     df[col] = pd.Categorical(df[col])
   return df
 
+def hierarchical_clustering_with_plotly(df, linkage_method):
+
+    df_encoded = df.apply(lambda x: pd.factorize(x)[0])
+
+    Z = linkage(df_encoded, method=linkage_method)
+
+    fig = ff.create_dendrogram(df_encoded, linkagefun=lambda x: Z, orientation='bottom')
+    fig.update_layout(width=800, height=500)
+    st.plotly_chart(fig)
+
+    num_clusters = int(input("Enter the desired number of clusters: "))
+
+    clusters = fcluster(Z, num_clusters, criterion='maxclust')
+
+    df['Cluster'] = clusters
+
+    return df
+
 empty_col1, main_col, empty_col2 = st.columns([1.6, 2.8, 1.6])
 
 with main_col:
@@ -808,9 +826,17 @@ if uploaded_file:
 
             matching_cols1 = []
             for i in multi_list:
-                matching_cols1 += [col for col in df.columns if is_matching_pattern(col, var1)]
+                matching_cols1 += [col for col in df.columns if is_matching_pattern(col, i)]
 
             df_clean = process_dataframe(df[single_list + matching_cols1])
+            st.subheader("Selected Table")
+            st.dataframe(df_clean)
+
+            linkage_method = st.sidebar.selectbox("Select the Linkage Method of Segmentation Analysis:", ["Hierarchical Clustering"])
+            
+            df_cluster = hierarchical_clustering_with_plotly(df_clean, linkage_method)
+
+            st.subheader("Cluster Table")
             st.dataframe(df_clean)