为具有空单元格的Python-dataframe的所有列生成单独的散点图

Question

我尝试自动绘制大型数据帧的相关性图。目标是将每一列与另一列在散点中绘制，并通过一条回归线。每一列代表一个不同的变量，一列可能有空单元格、整数和字符串值(尝试代码和工作示例如下)

示例代码：

Age     Height   Weight  Sex
21      180      54      M
56      171      65      V
23      NaN      84      V
NaN     195      71      M
42      165      67      V
84      167      93      M
12      NaN      88      M
31      152      73      V
NaN     184      NaN     V


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

df_subset = pd.DataFrame({"Age": [21,56,23,np.nan,42,84,12,31,np.nan], "Height": 
[180,171,np.nan,195,165,167,np.nan,152,184], "Weight": [54,65,84,71,67,93,88,73,np.nan], "Sex": 
['M','V','V','M','V','M','M','V','V']})

print(df_subset)
col_choice = ["Age", "Height", "Weight"]

for pos1, axis1 in enumerate(col_choice):   # Pick a first col
    for pos2, axis2 in enumerate(col_choice[pos1+1:]):   # Pick a later col
        plt.scatter(df_subset.loc[:,axis1], df_subset.loc[:,axis2]) #scatter plot
        a, b = np.polyfit(df_subset.loc[:,axis1], df_subset.loc[:,axis2], 1) #determining parameters for regression line
        x = df_subset.loc[:,axis1]
        plt.plot(x, a*x + b) #regression line on scatter-plot
        plt.show()

Answer 1

解决方案：

    import pandas as pd

    import numpy as np

    import matplotlib.pyplot as plt



    df_subset = pd.DataFrame({"Age": [21,56,23,np.nan,42,84,12,31,np.nan], "Height": 
[180,171,np.nan,195,165,167,np.nan,152,184], "Weight": [54,65,84,71,67,93,88,73,np.nan], "Sex": 
['M','V','V','M','V','M','M','V','V']})

 
    print(df_subset)



    col_choice = ["Age", "Height", "Weight"]

    for pos1, axis1 in enumerate(col_choice):   # Pick a first col
    
        for pos2, axis2 in enumerate(col_choice[pos1+1:]):   # Pick a later col
        
            df = df_subset[[axis1,axis2]].dropna()
        
            print(df)
        
            plt.scatter(df.iloc[:,0], df.iloc[:,1]) #scatter plot
        
            a, b = np.polyfit(df.iloc[:,0], df.iloc[:,1], 1) #determining parameters for regression line
        
            x = df.iloc[:,0]
        
            plt.plot(x, a*x + b) #regression line on scatter-plot
        
            plt.show()