数据集与探索性数据分析

1import numpy as np
2import pandas as pd
3import seaborn as sns
4sns.set(font_scale=1.3) # setting font size for the whole notebook
5sns.set_style("whitegrid") # setting the style
6import matplotlib.pyplot as plt

1df=pd.read_csv('Breast_Cancer_Diagnostic_Wisconsin.csv')
2print(df.info())

1print(df.describe())

1print(df.target.value_counts())

1# figure and the axes 
2fig1,ax=plt.subplots(nrows=1,ncols=2,figsize=(15,6),sharey=True)
3# groupby
4grouped_data=df.groupby("target")
5# plots
6grouped_data['mean radius'].hist(alpha=0.6,ax=ax[0])
7grouped_data['mean texture'].hist(alpha=0.6,ax=ax[1])
8# setting labels, title, and so
9ax[0].set_title("Mean radius distribution by class")
10ax[0].set_xlabel("Mean radius")
11ax[0].set_ylabel("Frequency")
12ax[1].set_title("Mean texture distribution by class")
13ax[1].set_xlabel("Mean texture")
14fig1.legend(["Malignant","Benign"],bbox_to_anchor=(0.57,0.92))

1import math
2
3nrows=6
4ncols=5
5fig2, ax = plt.subplots(nrows=nrows,ncols=ncols,figsize=(36,18))
6for i,col_ in enumerate(df.columns[:-1]): # Omitting target column
7    ax_ = ax[math.floor(i/ncols),i%ncols] # grabbing axis -- A nice trick to avoid nested loops!
8    df.groupby("target")[col_].hist(ax=ax_, alpha=.6) # plot
9    ax_.set_xlabel(col_)
10    ax_.set_ylabel("Frequency")
11    ax_.legend(["Malignant","Benign"])