analyze_visualisation.py

import os
import seaborn as sns
import pandas as pd
import matplotlib.pyplot as plt
import plotly.express as px


def make_and_save_pariplot(whole_df ,columns_to_analyze, file_name):

    current_dir = os.path.dirname(os.path.abspath(__file__))
    parent_dir = os.path.dirname(current_dir)
    if not os.path.exists(os.path.join(parent_dir, 'not_in_repo', file_name)):
        plot = sns.pairplot(data=whole_df[columns_to_analyze + ['our_final_status']], hue="our_final_status", palette={ 0 : "green", 1:"red"}, plot_kws={"s": 3})
        plt.legend(title='status końcowy')
        plot.savefig(os.path.join(parent_dir, 'not_in_repo', file_name))

def make_and_save_heatmap(whole_df, col_to_heatmap, file_name):

    current_dir = os.path.dirname(os.path.abspath(__file__))
    parent_dir = os.path.dirname(current_dir)
    if not os.path.exists(os.path.join(parent_dir, 'not_in_repo', file_name)):
        col_to_heatmap.append('our_final_status')
        selected_data = whole_df[col_to_heatmap]
        corrmat = selected_data.corr()
        plt.figure(figsize=(10, 8))
        sns.heatmap(corrmat, annot=True,cbar=True, cmap='Purples', fmt='.1f', linewidths=0.5,annot_kws={'size': 9})
        plt.title('Correlation Heatmap')
        plt.savefig(os.path.join(parent_dir, 'not_in_repo', file_name))

def describe_our_data(whole_df):
    pd.set_option('display.max_columns', None)
    pd.set_option('display.max_rows', None)
    pd.set_option('display.float_format', lambda x: '%.2f' % x)
    print(whole_df.describe().T)


def analyze_data(whole_df):
    #function take random data from dataframe and make pairplots and heatmap for all data
    status_1_data = whole_df[whole_df['our_final_status'] == 1].sample(n=200, random_state= 50)
    status_2_data = whole_df[whole_df['our_final_status'] == 2].sample(n=200, random_state= 50)
    random_to_analyze = pd.concat([status_1_data, status_2_data], ignore_index=True)

    col_dgm = ['czas_fazy_1', 'czas_fazy_2', 'czas_fazy_3', 'max_predkosc', 'cisnienie_tloka', 'cisnienie_koncowe','nachdruck_hub', 
                  'anguss', 'temp_pieca', 'oni_temp_curr_f1', 'oni_temp_curr_f2', 'oni_temp_fore_f1', 'oni_temp_fore_f2', 'vds_air_pressure',
                    'vds_vac_hose1', 'vds_vac_hose2', 'vds_vac_tank', 'vds_vac_valve1', 'vds_vac_valve2', 'czas_taktu']
    col_flow = [f'flow_{n}' for n in range(1,29)]
    col_delay = [f'start_delay_{n}' for n in range(1,29)]
    col_temp = [f'temp_{n}' for n in range(1,29)]

    make_and_save_pariplot(random_to_analyze, col_dgm, 'normalized_dirst_20_normalized.png')
    make_and_save_pariplot(random_to_analyze, col_flow, 'normalized_flow_normalized.png')
    make_and_save_pariplot(random_to_analyze, col_delay, 'normalize_delay_pairplot_normalized.png')
    make_and_save_pariplot(random_to_analyze, col_temp, 'normalize_temp_pairplot_normalized.png')
    make_and_save_heatmap(whole_df, col_dgm, 'cor1_hm.png')
    make_and_save_heatmap(whole_df, col_flow, 'flow_heatmap.png')
    make_and_save_heatmap(whole_df, col_delay, 'delay_heatmap.png')
    make_and_save_heatmap(whole_df, col_temp, 'temp_heatmap.png')