app.py

import streamlit as st
import pandas as pd 
import pandas as pd
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
from sklearn.linear_model import LinearRegression
from sklearn.metrics import r2_score
import matplotlib.pyplot as plt
st.set_option('deprecation.showPyplotGlobalUse', False)
from streamlit_extras.metric_cards import style_metric_cards

 

#navicon and header
st.set_page_config(page_title="Dashboard", page_icon="📈", layout="wide")  


st.header(" MACHINE LEARNING WORKFLOW ")
st.image("images/logo2.webp",caption="")
st.write("MULTIPLE REGRESSION WITH  SSE, SE, SSR, SST, R2, ADJ[R2], RESIDUAL")
 # load CSS Style
with open('styles.css')as f:
    st.markdown(f"<style>{f.read()}</style>", unsafe_allow_html = True)

#logo
st.sidebar.image("images/logos.webp",caption="MULTI-VARIABLE REGRESSION")
st.sidebar.title("ADD NEW VALUE")

df = pd.read_excel('Book1.xlsx')
X = df[['Dependant', 'Wives']]
Y = df['Projects']


# Fit a linear regression model
model = LinearRegression()
model.fit(X, Y)

# Make predictions
predictions = model.predict(X)

# Predictions on the same data
y_pred = model.predict(X)

#Regression coefficients (Bo, B1, B2)
intercept = model.intercept_ #Bo
coefficients = model.coef_ #B1, B2

# Calculate R-squared Coefficient of determination
r2 = r2_score(Y, predictions)

# Calculate Adjusted R-squared
n = len(Y)
p = X.shape[1]
adjusted_r2 = 1 - (1 - r2) * (n - 1) / (n - p - 1)

# Calculate Sum Squared Error (SSE) and SSR
sse = np.sum((Y - predictions)**2)
ssr = np.sum((y_pred - np.mean(Y)) ** 2)


#regression line
with st.expander("REGRESSION COEFFICIENT"):
 col1,col2,col3=st.columns(3)
 col1.metric('INTERCEPT:',value= f'{intercept:.4f}',delta="(Bo)")
 col2.metric('B1 COEFFICIENT:',value= f'{coefficients[0]:.4f}',delta=" for X1 number of Dependant (B1)")
 col3.metric('B2 COEFFICIENT',value= f'{coefficients[1]:.4f}',delta=" for X2 number of Wives (B2):")
 style_metric_cards(background_color="#FFFFFF",border_left_color="#9900AD",border_color="#1f66bd",box_shadow="#F71938")



# Print R-squared, Adjusted R-squared, and SSE
with st.expander("MEASURE OF VARIATIONS"):
 col1,col2,col3=st.columns(3)
 
 col1.metric('R-SQUARED:',value= f'{r2:.4f}',delta="Coefficient of Determination")
 col2.metric('ADJUSTED R-SQUARED:',value= f'{adjusted_r2:.4f}',delta="Adj[R2]")
 col3.metric('SUM SQUARED ERROR (SSE):',value= f'{sse:.4f}',delta="Squared(Y-Y_pred)")
 style_metric_cards(background_color="#FFFFFF",border_left_color="#9900AD",border_color="#1f66bd",box_shadow="#F71938")

 
 # Print a table with predicted Y

with st.expander("PREDICTION TABLE"):
 result_df = pd.DataFrame({'Name':df['Name'],'No of Dependant':df['Dependant'], 'No of Wives': df['Wives'], 'Done Projects | Actual Y': Y, 'Y_predicted': predictions})
 # Add SSE and SSR to the DataFrame
 result_df['SSE'] = sse
 result_df['SSR'] = ssr
 st.dataframe(result_df,use_container_width=True)

 #download predicted csv

 df_download = result_df.to_csv(index=False).encode('utf-8')
 st.download_button(
    label="DOWNLOAD PREDICTED DATASET",
    data=df_download,
    key="download_dataframe.csv",
    file_name="my_dataframe.csv"
 )


with st.expander("RESIDUAL & LINE OF BEST FIT"):
 
 # Calculate residuals
 residuals = Y - predictions
 # Create a new DataFrame to store residuals
 residuals_df = pd.DataFrame({'Actual': Y, 'Predicted': predictions, 'Residuals': residuals})
 # Print the residuals DataFrame
 st.dataframe(residuals_df,use_container_width=True)

 col1, col2=st.columns(2)
 with col1:
  plt.scatter(Y, predictions)
  plt.plot([min(Y), max(Y)], [min(Y), max(Y)], '--k',color='red', label='Best Fit Line')  # Best fit line
  plt.xlabel('Actual Y | number of Projects')
  plt.ylabel('Predicted Y')
  plt.grid(True)
  plt.legend()
  st.pyplot()

import seaborn as sns
with col2:
 sns.displot(residuals,kind='kde',color='blue', fill=True, legend=True)
 sns.set_style("whitegrid")  # Set the style to whitegrid
 st.pyplot()


# User input for X1 and X2
with st.sidebar:
 with st.form("input_form",clear_on_submit=True):
  x1 = st.number_input("Enter Dependant",)
  x2 = st.number_input("Number of Wives",)
  submit_button = st.form_submit_button(label="Predict")


if submit_button:
  # Make predictions
  new_data = np.array([[x1, x2]])
  new_prediction = model.predict(new_data)
  # Display prediction
  with st.expander("OPEN"):
   st.write(f"<span style='font-size: 34px;color:green;'>Predicted Output: </span> <span style='font-size: 34px;'> {new_prediction}</span>", unsafe_allow_html=True)