在python sklearn中安排多个回归结果

Question

I have coefficients from running multiple regression.

I want to arrange the independent variables as column names, and then place coefficients as the first row, to be able to see which independent variable has which coefficient.

#assigning independent variables to x
x = df.drop("dependent var", axis = 1)
...
#creating a data frame with independent variables as headers
df_coef = pd.DataFrame(columns = x.columns)
#adding a row with a coefficients
df.loc[0] = x
#shifting index by 1
df.index = df.index + 1
#sorting by index 
df = df.sort()

I am getting an error:

ValueError: cannot set a row with mismatched columns

What am I doing wrong?

Answer 1

The first command line:

#assigning independent variables to x
x = df.drop("dependent var", axis = 1)

What you are doing here is copying your original dataframe, df, without the "dependent var" column. (it was, as suggested by the command, dropped from the dataframe)

Whatever columns are left at df are than being associated as headers to your new df:

#creating a data frame with independent variables as headers
df_coef = pd.DataFrame(columns = x.columns)

Afterwise, you try to add a row to your new df, at index position 0, with a single value:

#adding a row with a coefficients
df.loc[0] = x

This is most likely where your error occurs, the df most likely has more than one column, yet, you try to create a row with a singular value and hence get the error message:

ValueError: cannot set a row with mismatched columns

You might be using an IDE to code, if so, a good advice is to use a more friendly data visualization tool, such as anaconda/jupyter notebook. With it you can visualize data step by step, which would help avoid the mistake done at the first step.

Hope that helps,