inital commit

Prediction.py

@@ -0,0 +1,87 @@
+import streamlit as st
+import pickle
+import pandas as pd
+from tensorflow.keras.models import load_model
+import numpy as np
+# from sklearn.pipeline import make_pipeline
+# from sklearn.preprocessing import StandardScaler, OneHotEncoder
+# from sklearn.svm import SVC
+# from sklearn.linear_model import LogisticRegression
+# from sklearn.tree import DecisionTreeClassifier
+# from sklearn.ensemble import RandomForestClassifier
+
+
+# Load the Models
+
+with open('final_pipeline.pkl', 'rb') as file_1:
+  model_pipeline = pickle.load(file_1)
+
+model_ann = load_model('churn_model.h5')
+
+def run():
+    with st.form(key='form_prediksi'):
+        name = st.text_input('Nama', value='')
+        sex = st.radio('Kelamin', ('Perempuan', 'Laki-Laki'))
+        if sex=='Laki-Laki':
+            gender='M'
+        else: gender='F'
+        age= st.number_input('Umur', min_value=16, max_value=80, value=50, step=1)
+        regcat=st.selectbox('Kategori Daerah: ',('Village','Town', 'City'))
+        memcat=st.selectbox('Kategori Membership: ',('No Membership','Basic Membership', 'Gold Membership', 'Premium Membership','Platinum Membership'))
+        ref = st.radio('apakah bergabung melalui referal?', ('Yes', 'No'))
+        medium=st.selectbox('Medium Akses: ',('Smartphone','Desktop', 'Both'))
+        preferensi=st.selectbox('Preferensi Penawaran: ',('Credit/Debit Card Offers', 'Gift Vouchers/Coupons','Without Offers'))
+        internet=st.selectbox('Preferensi Penawaran: ',('Fiber_Optic', 'Wi-Fi', 'Mobile_Data'))
+        daylast= st.number_input('Hari dari login terakhir', min_value=0, max_value=100, value=50, step=1)
+        avgday= st.number_input('Waktu pemakaiwn rata rata', min_value=0, max_value=100, value=50, step=1)
+        avgtran= st.number_input('Rata rata jumlah transaksi', min_value=0, max_value=50000, value=10000, step=1)
+        avgfreq= st.number_input('Hari dari login terakhit', min_value=0, max_value=30, value=10, step=1)
+        point= st.number_input('Point dalam Wallet', min_value=0, max_value=2000, value=50, step=1)
+        diskon= st.radio('Pernah menggunakan diskon spesial?', ('Yes', 'No'))
+        offer= st.radio('offer aplication prefrence?', ('Yes', 'No'))
+        past=st.radio('Pernah komplain?', ('Yes', 'No'))
+        complain= st.selectbox('Preferensi Penawaran: ',('Not Applicable', 'Unsolved', 'Solved', 'No Information Available','Solved in Follow-up'))
+        feedback= st.selectbox('Preferensi Penawaran: ',('Too many ads', 'No reason specified', 'Reasonable Price','Quality Customer Care', 'Poor Website', 'Poor Customer Service','Poor Product Quality', 'User Friendly Website', 'Products always in Stock'))
+
+        submitted = st.form_submit_button('Predict')
+
+    data_inf = {
+        'age': age,
+        'gender': gender,
+        'region_category':regcat,
+        'membership_category':memcat,
+        'joined_through_referral':ref, 
+        'preferred_offer_types':preferensi,
+        'medium_of_operation':medium, 
+        'internet_option':internet, 
+        'days_since_last_login':daylast,
+        'avg_time_spent':avgday, 
+        'avg_transaction_value': avgtran, 
+        'avg_frequency_login_days':avgfreq,
+        'points_in_wallet':point, 
+        'used_special_discount':diskon,
+        'offer_application_preference':offer, 
+        'past_complaint':past, 
+        'complaint_status':complain,
+        'feedback': feedback
+    }
+
+
+
+    if submitted:
+        data_inf = pd.DataFrame([data_inf])
+        # Transform Inference-Set
+
+        data_inf_transform = model_pipeline.transform(data_inf)
+        data_inf_transform
+        y_pred_inf = model_ann.predict(data_inf_transform)
+        y_pred_inf = np.where(y_pred_inf >= 0.5, 1, 0)
+        value = y_pred_inf[0][0]
+        print(value)
+        if value==1:
+            result= "Pelanggan diprediksi akan Churn"
+        else: result= "Pelanggan diprediksi tidak akan Churn"
+        st.write(result)
+
+if __name__== '__main__':
+    run()

app.py

@@ -0,0 +1,10 @@
+import streamlit as st
+import Prediction
+import eda
+
+navigation = st.sidebar.selectbox('Pilih Halaman: ', ('EDA','Prediction'))
+
+if navigation == 'EDA':
+    eda.run()
+else:
+   Prediction.run()

churn_model.h5

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:689b2bfd7af593f5b2de3704a57a9377eb7cbccc4bfb7afeef5be38c7fb19707
+size 115392

eda.py

@@ -0,0 +1,88 @@
+import streamlit as st
+import pandas as pd
+import seaborn as sns 
+import matplotlib.pyplot as plt
+import plotly.express as px
+from PIL import Image
+
+st.set_page_config(
+   page_title= 'Prediksi Churn Pelanggan',
+   layout='wide',
+   initial_sidebar_state='expanded'
+)
+
+def run():
+    image = Image.open('image.png')
+    resized_image = image.resize((605, 212))
+    st.image(resized_image, caption='Churn')
+
+    st.title('Prediksi Churn Pelanggan')
+    df = pd.read_csv('https://raw.githubusercontent.com/Azrieldr/latihan/master/churn.csv')
+    st.dataframe(df)
+
+    data=df.dropna()
+
+    #menggroupkan churn berdasarkan membership
+    churnMember =data.groupby('membership_category')['churn_risk_score'].mean()*100
+
+    # buat dictionary untuk mapping nama index yang baru
+    new_index = {}
+    for index in churnMember.index:
+        new_index[index] = index.replace('Membership', '')
+
+    # rename index dengan dictionary new_index
+    churnMember = churnMember.rename(index=new_index)
+    churnMember = churnMember.sort_values()
+
+    # plot bar chart
+    plt.bar(churnMember.index, churnMember.values, color='#89cff0')
+
+    # tambahkan judul dan label sumbu
+    plt.title('Rata-Rata Churn Risk Score per Kategori Membership')
+    plt.xlabel('Kategori Membership')
+    plt.ylabel('Rata-Rata Churn Risk Score (%)')
+
+    # simpan plot ke dalam variabel fig
+    fig = plt.gcf()
+
+    # tampilkan plot pada Streamlit
+    st.pyplot(fig)
+
+    # menggroupkan churn berdasarkan Complaint
+    churnCom = data.groupby('past_complaint')['churn_risk_score'].mean()*100
+
+    # plot bar chart
+    fig, ax = plt.subplots()
+    ax.bar(churnCom.index, churnCom.values, color='#89cff0')
+
+    # tambahkan judul dan label sumbu
+    ax.set_title('Rata-Rata Churn Risk Score berdasarkan pernahnya complaint')
+    ax.set_xlabel('Past Complaint')
+    ax.set_ylabel('Rata-Rata Churn Risk Score (%)')
+
+    # menampilkan plot pada Streamlit
+    st.pyplot(fig)
+
+
+    # membuat dataframe copy dari dataframe awal
+    df1 = data.copy()
+
+    # mengelompokkan data pada kolom AGE menjadi 10 kelompok
+    df1['group'] = pd.cut(df1['age'], bins=8)
+
+    # menghitung nilai rata-rata pada kolom LUNG_CANCER untuk setiap kelompok
+    result = df1.groupby('group')['churn_risk_score'].mean()*100
+
+    # plot hasilnya menggunakan seaborn dengan barplot berwarna pink
+    sns.set_style('whitegrid')
+    ax = sns.barplot(x=result.index, y=result, color='pink')
+    ax.set(xlabel='AGE Group', ylabel='churn risk score (YES)')
+
+    # mengatur font size pada axis x
+    ax.tick_params(axis='x', labelsize=7.5)
+
+    # menampilkan plot pada streamlit
+    st.pyplot(plt)
+
+if __name__== '__main__':
+    run()

final_pipeline.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6b831d1fc631ad1d90ddec4c9f227480c17c875158fad80208f8ea54249d4102
+size 3104