eda.py

import streamlit as st
st.title("The EDA Page")
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
st.set_option('deprecation.showPyplotGlobalUse', False)
plt.style.use('fivethirtyeight')

#load and clean the data
data=pd.read_csv("data/data.csv")
data.drop('Unnamed: 0', axis=1, inplace=True)
zerofiller=lambda x:x.replace(0, x.median())
cols=data.columns[1:6]
data[cols]=data[cols].apply(zerofiller, 0)
# encode the categorical outcome variable
df=data.copy()
d={"Yes":1, 'No':0}
df['Outcome']=df['Outcome'].map(d)
def view_data(data):
    st.write(df.head(10))
    st.pyplot()

def histograms(df):
    df.hist()
    plt.tight_layout()
    st.pyplot()
    
def barplot(data, feature):
    print("Bar plot of the variable ", feature)
    plt.figure(figsize=(10,7))
    ax=sns.countplot(data=data, x=feature, color='green')
    for p in ax.patches:
        x=p.get_bbox().get_points()[:,0]
        y=p.get_bbox().get_points()[1,1]
        ax.annotate("{:.2g}%".format(100.*y/len(df)), (x.mean(), y), ha='center', va='bottom')
    st.pyplot()
    
def boxplot_histplot(data, feature, bins=None, figsize=(12,7)):
    print('Boxplot and Histplot for ', feature)
    fig, (ax_box, ax_hist)=plt.subplots(
    nrows=2,
    sharex=True,
    gridspec_kw = {"height_ratios":(0.25, 0.75)},
    figsize=figsize
    )
    sns.boxplot(data=data, x=feature, color='violet', ax=ax_box, showmeans=True)
    sns.histplotplot(data=data, x=feature, ax=ax_hist, bins=bins)   if bins else sns.histplot(data\
                            =data, x=feature, ax=ax_hist)  
    ax_hist.axvline(data[feature].mean(), color='green', linestyle='--')
    ax_hist.axvline(data[feature].median(), color='black', linestyle='-')
    st.pyplot()
    
st.sidebar.subheader("Choose the Plot")
plot=st.sidebar.selectbox('plot', ('Data', 'Histograms', 'Barchart', 'Boxplot_Scatterplot', 'Correlations'))

if st.sidebar.button('PLOT'):
    if plot=='Data':
        view_data(df)
    if plot=='Histograms':
        histograms(df)
    if plot=='Barchart':
        barplot(df, feature='Outcome')
    if plot=='Boxplot_Scatterplot':
        for col in df.select_dtypes(exclude='O').columns:
            boxplot_histplot(df, col)
    if plot=='Correlations':
        plt.figure(figsize=(12,7))
        sns.heatmap(df.corr(), cmap='Spectral', vmax=+1, vmin=-1, annot=True)
        st.pyplot()