helper.py

from sys import prefix
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
import streamlit as st
import datetime, pytz
import glob, os
from sklearn.preprocessing import LabelEncoder


excel_type =["vnd.ms-excel","vnd.openxmlformats-officedocument.spreadsheetml.sheet", "vnd.oasis.opendocument.spreadsheet", "vnd.oasis.opendocument.text"]


def data(data, file_type, seperator=None):

    if file_type == "csv":
        data = pd.read_csv(data)
    
    elif file_type in excel_type:
        data = pd.read_excel(data)
        st.sidebar.info("If you are using Excel file so there could be chance of getting minor error.")
    
    elif file_type == "plain":
        try:
            data = pd.read_table(data, sep=seperator)
        except ValueError:
            st.info("Enter delimiter")

    return data


def download_data(data, label):
    current_time = datetime.datetime.now(pytz.timezone('Asia/Kolkata'))
    current_time = "{}.{}-{}-{}".format(current_time.date(), current_time.hour, current_time.minute, current_time.second)
    export_data = st.download_button(
                        label="Download {} data as CSV".format(label),
                        data=data.to_csv(),
                        file_name='{}{}.csv'.format(label, current_time),
                        mime='text/csv',
                        help = "When You Click On Download Button You can download your {} CSV File".format(label)
                    )
    return export_data


def describe(data):
    global num_category, str_category
    num_category = [feature for feature in data.columns if data[feature].dtypes != "O"]
    str_category = [feature for feature in data.columns if data[feature].dtypes == "O"]
    column_with_null_values = data.columns[data.isnull().any()]
    return data.describe(), data.shape, data.columns, num_category, str_category, data.isnull().sum(),data.dtypes.astype("str"), data.nunique(), str_category, column_with_null_values


def outliers(data, num_category_outliers):
    plt.figure(figsize=(6,2))
    flierprops = dict(marker='o', markerfacecolor='purple', markersize=6,
                    linestyle='none', markeredgecolor='black')
    
    path_list = []
    for i in range(len(num_category_outliers)):
        
        column = num_category_outliers[i]
        plt.xlim(min(data[column]), max(data[column])) 
        plt.title("Checking Outliers for {} Column".format(column))
        plot = sns.boxplot(x=column, flierprops=flierprops, data=data)
        fig = plot.get_figure()

        path = 'plots/pic{}.png'.format(i)
        fig.savefig(path)
        path_list.append(path)

    return path_list


def label_encode(data,selected_name):
    le = LabelEncoder()
    arr = le.fit_transform(data[selected_name])
    df = pd.DataFrame(arr, columns = [selected_name])
    data[selected_name]=df
    return data

def drop_items(data, selected_name):
    droped = data.drop(selected_name, axis = 1)
    return droped


def filter_data(data, selected_column, selected_name):
    if selected_name == []:
        filtered_data = data
    else:
        filtered_data = data[~ data[selected_column].isin(selected_name)]
    return filtered_data


def num_filter_data(data, start_value, end_value, column, param):
    if param == "Delete data inside the range":
        if column in num_category:
            num_filtered_data = data[~data[column].isin(range(int(start_value), int(end_value)+1))]
    else:
        if column in num_category:
            num_filtered_data = data[data[column].isin(range(int(start_value), int(end_value)+1))]
    
    return num_filtered_data


def rename_columns(data, column_names):
    rename_column = data.rename(columns=column_names)
    return rename_column


def handling_missing_values(data, option_type, dict_value=None):
    if option_type == "Drop all null value rows":
        data = data.dropna()

    elif option_type == "Only Drop Rows that contanines all null values":
        data = data.dropna(how="all")
    
    elif option_type == "Filling in Missing Values":
        data = data.fillna(dict_value)
    
    return data


def clear_image_cache():
    removing_files = glob.glob('plots/*.png')
    for i in removing_files:
        os.remove(i)