1
2
3
4
5
6
7
8
import pandas as pd
import numpy as np
import seaborn as sns
import os
import matplotlib.pyplot as plt

plt.style.use('seaborn-colorblind')
%matplotlib inline

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
def get_dtypes(data, drop_col = []):
    # list columns kieu str, list columns kieu number, list all
    # data: dataframe
    # drop_col: bo mot so cot tuong ung.
    
    name_of_col = list(data.columns)
    num_var_list = []
    str_var_list = []
    all_var_list = []
    
    str_var_list = name_of_col.copy()
    
    for var in name_of_col:
        # check if column belongs to numeric type
        if data[var].dtypes in (np.int64, np.int32, np.float64, np.float32, np.double):
            str_var_list.remove(var)
            num_var_list.append(var)
    # drop the omit column from list
    for var in drop_col:
        if var in str_val_list:
            str_val_list.remove(var)
        if var in num_var_list:
            num_var_list.remove(var)
    
    all_var_list.extend(str_var_list)
    all_var_list.extend(num_var_list)
    
    return str_var_list, num_var_list, all_var_list

def describe(data, output_path = None):
    res = data.describe(include = 'all')
    if output_path is not None:
        output = os.path.join(output_path, 'describe.csv')
        res.to_csv(output)
        print('result saved at: ' + str(output))
    return res
def discrete_var_barplot(x, y, data, output_path = None):
    # ve bieu do bien roi rac x doi voi ham muc tieu y
    # default: bars show mean value of y
    plt.figure(figsize = (15, 10))
    sns.barplot(x = x, y = y, data = data)
    
    if output_path is not None:
        output = os.path.join(output_path, 'Barplot_' + str(x) + '_' + str(y) + '.png')
        plt.savefig(output)
        print('image saved at', str(output))
def discrete_var_countplot(x, data, output_path = None):
    # count value x 
    plt.figure(figsize = (15, 10))
    sns.countplot(x = x, data = data)
    if output_path is not None:
        output = os.path.join(output_path, 'Countplot_' + str(x) + '.png')
        plt.savefig(output)
        print('Image saved at', str(output))

def discrete_var_boxplot(x, y, data, output_path = None):
    # ve bieu do bien roi rac x doi voi ham muc tieu y
    # default: bars show mean value of y
    plt.figure(figsize = (15, 10))
    sns.boxplot(x = x, y = y, data = data)
    
    if output_path is not None:
        output = os.path.join(output_path, 'Boxplot_' + str(x) + '_' + str(y) + '.png')
        plt.savefig(output)
        print('image saved at', str(output))
def continuous_var_histplot(x, output_path = None, bins = None):
    # draw the histogram of a continuous variable x
    
    plt.figure(figsize = (15, 10))
    sns.histplot(data = x)
    if output_path is not None:
        output = os.path.join(output_path, 'Histplot_' + str(x.name) + '.png')
        plt.savefig(output)
        print('Image saved at', str(output))
def scatter_plot(x, y, data, output_path = None):
    plt.figure(figsize = (15, 10))
    sns.scatterplot(x = x, y = y, data = data)
    if output_path is not None:
        output = os.path.join(output_path, 'Scatter_plot_' + str(x.name) + '_' + str(y.name) + '.png')
        plt.savefig(output)
        print('Image saved at', str(output))
def correlation_plot(data, output_path = None):
    corrmat = data.corr() # correlation matrix
    fig, ax = plt.subplots()
    fig.set_size_inches(11, 11)
    sns.heatmap(corrmat, cmap = 'YlGnBu', linewidths = 0.5, annot = True)
    if output_path is not None:
        output = os.path.join(output_path, 'Corr_plot' + '.png')
        plt.savefig(output)
        print('Image saved at', str(output))
def heatmap(data, output_path = None, fmt = 'd'):
    fig, ax = plt.subplots()
    fig.set_size_inches(11, 11)
    sns.heatmap(data, cmap = 'YlGnBu', linewidths = .5, annot = True, fmt = fmt)
    if output_path is not None:
        output = os.path.join(output_path, 'Heatmap' + '.png')
        plt.savefig(output)
        print('Image saved at', str(output))

1
2
3
4
5
6
path = 'https://raw.githubusercontent.com/dophuchao/dophuchao.github.io/master/data/titanic.csv'
use_cols = [
    'Pclass', 'Sex', 'Age', 'Fare', 'SibSp', 'Survived'
]
data = pd.read_csv(path, usecols = use_cols)
data.head(3

1
2
3
4
str_var_list, num_var_list, all_var_list = get_dtypes(data)
print(str_var_list)
print(num_var_list)
print(all_var_list)

1
describe(data = data, output_path = r'./output/')

1
discrete_var_barplot(x = 'Pclass', y ='Survived', data = data, output_path = r'./output/')

1
discrete_var_countplot(x = 'Pclass', data = data, output_path = r'./output/')

1
2
3
# chu y tham so:
# min, 1st Quartile, Median, Mean, 3rd Quartile, Max
discrete_var_boxplot(x = 'Pclass', y ='Fare', data = data, output_path = r'./output/')

1
continuous_var_histplot(x = data['Fare'], output_path = r'./output/')

1
2
# muc do tuong quan cua x, y
scatter_plot(x = data.Fare, y = data.Pclass, data = data, output_path = r'./output/')

1
correlation_plot(data = data, output_path = r'./output/')

1
2
3
4
flights = sns.load_dataset('flights')
print(flights.head(3))
flights = flights.pivot('month', 'year', 'passengers')
heatmap(data = flights, output_path = r'./output/')