UCI數據集iris數據簡單的可視化

數據集官網下載；

jupyter notebook 實現；

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



fname = 'E:\\pythonwork\\project\\Deeplearning\\Task\\data\\iris.data'
with open(fname, 'r+', encoding='utf-8') as f:
    s = [i[:-1].split(',') for i in f.readlines()]
# 讀取TXT，逗號為分隔符

# pandas讀取數據 樣本數為各50個
names=['slength','swidth','plength','pwidth','name']
iris = pd.DataFrame(data=s,  columns=names)
# 刪除一個莫名其妙的空行：
iris.dropna(axis=0, how='any', inplace=True)
# 有三種類別：
seto = iris.iloc[0:50,:]
vers = iris.iloc[50:100,:]
virg = iris.iloc[100:150,:]
seto.shape
vers.shape
# 統計每個品種有多少個樣本
iris['name'].value_counts()
# 字符串類型的數據變成float（否則不能畫圖）
iris.iloc[:,:4]=iris.iloc[:,:4].astype('float')
# 畫出slength和swidth的關系圖
plt.scatter(x=iris['slength'],y=iris['swidth'])
plt.show()

 

 

#-------------------
# 按顏色不同分類 畫圖
plt.scatter(x=seto['slength'],y=seto['swidth'],color='red')
plt.scatter(x=vers['slength'],y=seto['swidth'],color='blue',marker="+")
plt.scatter(x=virg['slength'],y=seto['swidth'],color='green',marker='*')
plt.xlabel('s length')
plt.ylabel('s width')
plt.show()