revise plot using plotly express

app.py

@@ -14,6 +14,7 @@ with st.sidebar:
         [
             "Distribution",
             "Classification",
+            "Cluster"
         ],
         icons=["bar-chart", "link-45deg", "code-square"],
         menu_icon="cast",
@@ -24,3 +25,5 @@ if selected == "Distribution":
     eda.distribution()
 elif selected == "Classification":
     prediction.predict()
+elif selected == "Cluster":
+    prediction.cluster()

eda.py

@@ -1,8 +1,6 @@
 import streamlit as st
 import pandas as pd
 import numpy as np
-import seaborn as sns
-import matplotlib.pyplot as plt
 import plotly.express as px
 
 bank_df = pd.read_csv('./csv/BankChurners.csv')
@@ -10,8 +8,6 @@ bank_df.drop(columns=["CLIENTNUM",
                       "Naive_Bayes_Classifier_Attrition_Flag_Card_Category_Contacts_Count_12_mon_Dependent_count_Education_Level_Months_Inactive_12_mon_1",
                       "Naive_Bayes_Classifier_Attrition_Flag_Card_Category_Contacts_Count_12_mon_Dependent_count_Education_Level_Months_Inactive_12_mon_2"], inplace=True)
 
-sns.set(style="whitegrid")
-palette=["teal", "darkblue"]
 num_col = bank_df.select_dtypes(include=np.number).columns.tolist()
 cat_col = bank_df.select_dtypes(include=object).columns.tolist()
 cat_col.remove("Attrition_Flag")
@@ -24,20 +20,17 @@ st.set_page_config(
 
 def distribution():
     # distribution plot
-    
     st.header("Data Distribution")
     
     attr_plot('Attrition_Flag')
     
     col1, col2 = st.columns(2)
     
-    numerik = col1.selectbox(label="Select Features",
-                   options=num_col)
+    numerik = col1.selectbox(label="Select Features", options=num_col)
     
     hist_plot(numerik, col1)
     
-    kategorik = col2.selectbox(label="Select Features",
-                   options=cat_col)
+    kategorik = col2.selectbox(label="Select Features", options=cat_col)
     
     count_plot(kategorik, col2)
     
@@ -46,19 +39,17 @@ def distribution():
                 ''')
 
 def attr_plot(column):
-    fig = plt.figure(figsize=(15,5))
-    sns.countplot(data=bank_df, y=column, palette=palette, alpha=0.7)
-    st.pyplot(fig)
+    fig = px.histogram(bank_df, y=column, color="Attrition_Flag", title=f'Distribution of {column}')
+    fig.update_layout(width=1200)
+    st.plotly_chart(fig, use_container_width=True)
 
 def hist_plot(column, loc):
-    fig = plt.figure(figsize=(15,6))
-    sns.histplot(data=bank_df, x=column, kde=True, bins=50, palette=palette, hue="Attrition_Flag")
-    loc.pyplot(fig)
-    
-def count_plot(column,loc):
-    fig = plt.figure(figsize=(15,6))
-    sns.countplot(data=bank_df, y=column, palette=palette, hue="Attrition_Flag", order=bank_df[column].value_counts().index, alpha=0.7)
-    loc.pyplot(fig)
-    
+    fig = px.histogram(bank_df, x=column, color="Attrition_Flag", title=f'Histogram of {column}')
+    loc.plotly_chart(fig)
+
+def count_plot(column, loc):
+    fig = px.bar(bank_df, y=column, color="Attrition_Flag", title=f'Count Plot of {column}', orientation='h')
+    loc.plotly_chart(fig)
+
 if __name__ == "__main__":
     distribution()

prediction.py

@@ -1,11 +1,7 @@
 import streamlit as st
 import pandas as pd
-import numpy as np
 import pickle
-import matplotlib.pyplot as plt
-from PIL import Image
-from urllib import request
-from io import BytesIO
+import plotly.express as px
 
 
 def predict():
@@ -103,6 +99,26 @@ def predict():
                         """
         st.markdown(result_html.format(pred_inf=pred_inf, cluster_inf=cluster_inf, color=color, step=recommendation), unsafe_allow_html=True)
 
-
+def cluster():
+    clusters = pd.read_csv('./csv/Cluster.csv')
+    bank_df_pca = pd.read_csv('./csv/BankPCA.csv')
+    
+    colors = {0: 'navy', 1: 'teal'}
+    names = {0: 'High Spent Amount (>4K), High Usage Frequency', 
+             1: 'Low Spent Amount (<4K), Low Usage Frequency'}
+    
+    bank_df_pca['color'] = bank_df_pca['label'].map(colors)
+    bank_df_pca['name'] = bank_df_pca['label'].map(names)
+    
+    fig = px.scatter(bank_df_pca, x='x', y='y', color='name', hover_name='name',
+                     title='Churn Customer Clustering', width=800, height=400, )
+    
+    fig.update_traces(marker=dict(size=5))
+    fig.update_layout(showlegend=True)
+    
+    fig.update_layout(height=600)
+    st.plotly_chart(fig, use_container_width=True)
+    
+    
 if __name__ == "__main__":
     predict()