streamlit

22-06-11-Streamlit/Procfile

@@ -0,0 +1 @@
+web: sh setup.sh && streamlit run app.py

22-06-11-Streamlit/README.md

@@ -0,0 +1,39 @@
+Live at https://obscure-reef-27133.herokuapp.com/
+
+create new environment
+`python -m venv env`
+
+activate env
+linux `source ./env/bin/activate`
+windows `env\Scripts\activate`
+
+`pip install streamlit`
+`pip install sklearn`
+
+create app.py
+
+Run file
+`streamlit run app.py`
+
+###Create Input fields
+(Streamlit docs)[https://docs.streamlit.io/library/api-reference/widgets/st.number_input]
+
+###Deploy in Heroku
+https://devcenter.heroku.com/articles/getting-started-with-python
+
+`pip freeze > requirements.txt`
+
+`git init`
+`git checkout -b main`
+`heroku create`
+
+Whenever file is modified run this command
+`git add file_name`
+`git commit -m 'init'`
+`git push heroku main`
+
+###Procfile
+command to run, this is python app
+
+###runtime.txt
+python version to use

22-06-11-Streamlit/app.py

@@ -0,0 +1,93 @@
+import streamlit as st
+import pandas as pd
+import numpy as np
+from sklearn.compose import ColumnTransformer
+from sklearn.impute import SimpleImputer
+from sklearn.preprocessing import OneHotEncoder, StandardScaler
+from sklearn.pipeline import Pipeline
+from sklearn.linear_model import LogisticRegression
+
+from datetime import datetime
+
+st.header("Bank Loan Prediction")
+st.image('https://miro.medium.com/max/640/1*UC0sy0bENl-DLPy3jmXNag.jpeg')
+
+
+@st.cache
+def get_data():
+    X = pd.read_csv("trainDataset.csv")
+    X.drop('Loan_ID', axis = 1, inplace = True)
+
+    Y = X['Loan_Status']
+    X = X.drop('Loan_Status', axis=1)
+
+    # Data Cleaning
+    X['Dependents'].replace({'3+': 5}, inplace = True)
+    X['Dependents'] = pd.to_numeric(X['Dependents'])
+    Y = np.where(Y == 'Y', 1, 0)
+    return X, Y
+
+X, Y = get_data()
+
+coltrans = ColumnTransformer([
+      ('imputer-mod_onehot', Pipeline([
+                                ('imputer-mod', SimpleImputer(strategy='most_frequent')),
+                                ('onehot', OneHotEncoder(drop='first')),
+                              ]), ['Gender', 'Married', 'Self_Employed']),
+      ('imputer-mod', SimpleImputer(strategy='most_frequent'), ['Credit_History']),
+      ('imputer-mean_scaling', Pipeline([
+                                ('imputer-mean', SimpleImputer(strategy='mean')),
+                                ('scaling', StandardScaler()),
+                              ]), ['Dependents', 'LoanAmount', 'Loan_Amount_Term']),
+      ('onehot', OneHotEncoder(drop='first'), ['Education', ]),
+      ('scaling', StandardScaler(), ['ApplicantIncome', 'CoapplicantIncome', ]),
+])
+
+X = coltrans.fit_transform(X)
+
+@st.cache
+def train_model():
+    model = LogisticRegression().fit(X, Y)
+    return model
+
+model = train_model()
+
+data = {}
+
+data['Gender'] = st.selectbox('Gender', ('Male', 'Female'))
+
+data['Married'] = st.selectbox('Married', ('Yes', 'No'))
+
+data['Dependents'] = st.number_input('How many people dependents on you', min_value=0)
+if data['Dependents'] > 3:
+    data['Dependents'] = 5
+
+data['Education'] = st.selectbox('Are you Graduated', ('Graduate', 'Not Graduate'))
+data['Self_Employed'] = st.selectbox('Bussiness', ('Yes', 'No'))
+data['ApplicantIncome'] = st.number_input("Your Income", min_value=0)
+data['CoapplicantIncome'] = st.number_input("Coapplicant Income", min_value=0)
+data['LoanAmount'] = st.number_input("Loan Amount", min_value=0)
+
+data['Loan_Amount_Term'] = st.number_input('Loan_Amount_Term', min_value=0)
+data['Property_Area'] = st.selectbox('Property_Area', ['Rural', 'Urban', 'Semiurban'])
+
+
+data['Credit_History'] = st.selectbox('Transaction Frequency', ['Good', 'Average']) # format_func
+if data['Credit_History'] == 'Good':
+    data['Credit_History'] = 1
+else:
+    data['Credit_History'] = 0
+
+
+features = pd.DataFrame(data, index=[0])
+st.write(features)
+
+features = coltrans.transform(features)
+
+pred = model.predict(features)
+
+if pred[0] == 1:
+    st.write('Congrats you will get loan')
+else:
+    st.write('You have very less chance of getting loan')    
+

22-06-11-Streamlit/requirements.txt

@@ -0,0 +1,2 @@
+scikit-learn==1.1.1
+streamlit==1.10.0

22-06-11-Streamlit/runtime.txt

@@ -0,0 +1 @@
+python-3.8.10

22-06-11-Streamlit/setup.sh

@@ -0,0 +1,3 @@
+mkdir -p ~/.streamlit
+echo "[server]" >> ~/.streamlit/config.toml
+echo "port = $PORT" >> ~/.streamlit/config.toml