浙大疏锦行作业01020304050607080910111213141516171819202122232425262728293031323334353637383940414243444546474849505152535455565758596061626364656667686970importpandas as pdheartdfpd.read_csv(Python打卡训练营\Python打卡训练营\python60-days-challenge-master\heart.csv)# 先看数据长什么样print(heartdf.head())print(heartdf.shape)print(heartdf.columns)print(heartdf.dtypes)print(heartdf.info())# 体检看看有无异常值print(heartdf.describe(percentiles[.25, .5, .75, .9]))# 检查缺失值有则进行清洗、填补print(heartdf.isnull().sum())missing_ratioheartdf.isnull().sum()/len(heartdf)*100print(missing_ratio)# 对于非分类的特征画图查看是否有异常值importmatplotlib.pyplot as pltimportseaborn as snsimportmathcategorical_features[sex,cp,fbs,restecg,exang,slope,ca,thal]labeltargetcontinuous_features[colforcolinheartdf.columnsifcolnotincategorical_featuresandcol !label]n_cols3n_rowsmath.ceil(len(continuous_features)/n_cols)fig, axesplt.subplots(n_rows,n_cols,figsize(15,4*n_rows))# 转成一维方便遍历axesaxes.flatten()fori, colinenumerate(continuous_features):sns.boxplot(xheartdf[col],axaxes[i])axes[i].set_title(fBoxplot of {col})# 如果最后一行没有填满把多余子图删掉forjinrange(len(continuous_features),len(axes)):fig.delaxes(axes[j])plt.tight_layout()plt.show()# 后续还要做标准化防止数据泄露划分数据集时间序列还需要做滑动窗口