python 读取csv文件并输出柱状图(以iris数据集为例)

被半自动的直方图法气晕

 

 

import pandas as pd
import numpy as np
import matplotlib.pylab as plt

data = pd.read_csv("iris.csv")

# 获取所有列，并存入一个数组中

ori_data = np.array(data)

#存储三种类别的list
list1 = []
list2 = []
list3 = []
max_num = 40
input_num = [0, 0, 0]
input_num1 = 0
data_num = 150

for i in range(data_num):
    if (ori_data[i][5] == 'setosa') & (input_num[0] < max_num):
        list1.append(ori_data[i])
        input_num[0] = input_num[0] + 1
    if (ori_data[i][5] == 'versicolor') & (input_num[1] < max_num):
        list2.append(ori_data[i])
        input_num[1] = input_num[1] + 1
    if (ori_data[i][5] == 'virginica') & (input_num[2] < max_num):
        list3.append(ori_data[i])
        input_num[2] = input_num[2] + 1

#提取iris三个种类的数据集为四个部分并画图
def mission1(list,class_name):
    data1 = [[], [], [], []]
    #数据名
    data_name = ["Sepal.Length","Sepal.Width","Petal.Length","Petal.Width"]
    #提取4种数据
    for i in range(len(list)):
        for j in range(4):
            data1[j].append(list[i][j+1])#数据特征

    # arange(a,b,c)确定直方图x轴的范围及间距，a为最小值，b为最大值，c为间距
    # 用plt.hist(a,b)绘制，a为数据，b为直方图的特性
    for i in range(4):
        print("序列", i)
        plt.subplot(2, 2, i+1)
        data = data1[i]
        # print(data)
        # 计算bin
        a = int(max(data)*10/10)
        b = int(min(data)*10/10)
        c = a - b
        bin = c + 1

        #直接输出每个区间的值
        num_count = [0, 0, 0, 0, 0, 0, 0, 0, 0, 0] #......代表统计范围是0-9
        for j in range(len(data)):
            flag = 0
            # print("输入：",data[j], " 输出：", int(data[j]*10/10))
            if int(data[j]*10/10) == b + flag:
                num_count[b + flag] = num_count[b + flag] + 1
            if int(data[j]*10/10) != b + flag:
                for h in range(bin):
                    flag = flag + 1
                    if int(data[j]*10/10) == b + flag:
                        num_count[b + flag] = num_count[b + flag] + 1
                        break

        #输出
        x = []
        y = []
        for s in range(len(num_count)):
            if num_count[s] != 0:
                # print("区间: ", s, " 数量： ", num_count[s])
                x.append(s)
                y.append(num_count[s])
        # print('max = ', max(data), 'min = ', min(data), 'minus = ', max(data) - min(data),'a = ', a, 'b = ', b, 'c = ', c, ' bin = ', bin)
        plt.bar(x,y,align='center',color="c",tick_label=x)
        # plt.hist(data, bin, (int(min(data)*10/10), int(max(data)*10/10)),rwidth = 0.6)
        plt.xlabel('x')
        plt.ylabel('num')
        
        for X,Y in zip(x,y):
            plt.text(X+0.05,Y+0.05,'%d' %Y, ha='center',va='bottom')

        # plt.xticks(np.arange(int(min(data)*10/10), int(max(data)*10/10), 1))
        plt.yticks(np.arange(0, 40, 2)) 
        title = str(class_name + ' ' + data_name[i])
        plt.title(title)

        #计算概率估计表
        for t in range(len(x)):
            result1 = y[t] / max_num
            print("范围", x[t], "Na/N", result1)


    plt.show()

mission1(list1,'setosa')
mission1(list2,'versicolor')
mission1(list3,'virginica')