python画图 箱线图 累积图 密度曲线图 数据点 标签 图例 合图

图片

话不多数，先上一波效果图

代码

# -*- coding: utf-8 -*-
"""
Created on Fri Jan 22 08:33:24 2021

@author: dujidan
"""

import os
import numpy as np
import seaborn as sns
import matplotlib.pyplot as plt

plt.style.use("ggplot")
# 解决中文显示问题
plt.rcParams['font.sans-serif'] = ['SimHei']
plt.rcParams['axes.unicode_minus'] = False


def read_column_float_file(input_file, start_row=0, n_col='all'):
    out_list = []
    with open(input_file, encoding='utf-8') as f_in:
        data_list = f_in.read().strip('\n').split('\n')
        if n_col == 'all':
            for line in data_list[start_row:]:
                out_list.append([float(j) for j in line.split('\t') if j != ''])
        else:
            for line in data_list[start_row:]:
                col_value = line.split('\t')[n_col]
                if col_value != '':
                    out_list.append(float(col_value))
    return out_list, data_list[0].split('\t')


# 箱线图，合图 2*2比较
def boxplot_2D_inOne(list_1, list_2, out_file='./boxplot_2D.png', suptitle_name='Box plot ', nrow=5, ncol=8):
    """
    list_1, list_2: []  1D;  len(A) = len(B)
    out_file: /home/test.png
    """
    plt.figure(figsize=(50, 30))
    for i in range(len(list_1)):
        plt.subplot(nrow, ncol, i+1)
        plt.boxplot((list_1[i], list_2[i]), labels=('before', 'after'))
        plt.legend()
    plt.suptitle(suptitle_name, fontsize=60)
    plt.savefig(out_file, dpi=500)


# 密度曲线图，合图 2*2比较
def distplot_2D_inOne(list_1, list_2, out_file='./doxplot_2D.png', suptitle_name='Dist plot', nrow=5, ncol=8):
    plt.figure(figsize=(50, 30))
    for i in range(len(list_1)):
        plt.subplot(nrow, ncol, i+1)
        sns.distplot(list_1[i], kde=True, hist=True, label='before')
        sns.distplot(list_2[i], kde=True, hist=True, label='after')
        plt.legend()
        plt.xlabel('深度')
        plt.ylabel('密度曲线')
    plt.suptitle(suptitle_name, fontsize=60)
    plt.savefig(out_file, dpi=500)


def read_float_file(input_file, start_row=0):
    out_list, index_list = [], []
    with open(input_file, encoding='utf-8') as f_in:
        data_list = f_in.read().strip().split('\n')
        index_list = data_list[0]
        for line in data_list[start_row:]:
            out_list.append([float(j) for j in line.split('\t') if j != ''])
    return out_list, index_list


def plot_percent_line(all_list, index_list):
    """
    all_list：2D list
    index_list：图例名称列表
    """
    fig, ax = plt.subplots()
    label_count = -1
    x_ticks = np.arange(0, 1, 0.1)
    for a_list in all_list:
        label_count += 1
        ret_list = []
        for set_num in x_ticks:
            count = 0
            for i in a_list:
                if i >= set_num:
                    count += 1
            ret_list.append(count/len(a_list))

        ax.plot(x_ticks, ret_list, 'o-', label=index_list[label_count])
        plt.xticks(x_ticks)
        for i, txt in enumerate(ret_list):
            ax.annotate(txt, (x_ticks[i], ret_list[i]))
            print(x_ticks[i], ret_list[i])
    plt.legend()
    plt.title('score 累计趋势')
    plt.xlabel(' 得分')
    plt.ylabel('> 得分 的位点占比')
    plt.savefig('得分统计分布-累计图.png', dpi=500)


if __name__ == "__main__":
    depth_file_20 = '202012_all_depth.xls'
    depth_file_21 = '202101_all_depth.xls'
    list_1, __ = read_column_float_file(depth_file_20, start_row=1, n_col='all')
    list_2, __ = read_column_float_file(depth_file_21, start_row=1, n_col='all')
    distplot_2D_inOne(list_1, list_2, out_file='./doxplot_2D.png', suptitle_name='Dist plot', nrow=5, ncol=8)
    boxplot_2D_inOne(list_1, list_2, out_file='./boxplot_2D.png', suptitle_name='Box plot ', nrow=5, ncol=8)

    # score
    score_file = 'score.xls'
    a_list, __ = read_column_float_file(score_file, start_row=1, n_col=0)
    b_list, __ = read_column_float_file(score_file, start_row=1, n_col=1)
    c_list, __ = read_column_float_file(score_file, start_row=1, n_col=2)
    d_list, index_list = read_column_float_file(score_file, start_row=1, n_col=3)


    plt.boxplot((a_list, b_list, c_list, d_list), labels=(index_list))
    plt.title('得分统计分布-箱线图')
    plt.savefig('得分统计分布-箱线图.png', dpi=500)

    plt.title('得分统计分布-密度图')
    sns.distplot(a_list, label=index_list[0])
    sns.distplot(b_list, label=index_list[1])
    sns.distplot(c_list, label=index_list[2])
    sns.distplot(d_list, label=index_list[3])
    plt.legend()
    plt.savefig('得分统计分布-密度图.png', dpi=500)

    all_list = [a_list, b_list, c_list, d_list]
    plot_percent_line(all_list, index_list