如何在 python pandas 的 for 循环中将分类数据转换为数值数据

Question

I have a categorical data framework and I want to convert it into numerical data, I have more than 50 columns so I want to run.repalce command in a loop.

replace_map =  {'w': 4, '+': 5, '.': 6, 'g': 7}

and I have written code which iterates over columns

for column in df1_replace.columns[1:76]:
# Select column contents by column name using [] operator
columnSeriesObj = df1_replace[column]
print('Colunm Name : ', column)
print('Column Contents : ', columnSeriesObj.values)

Answer 1

Here is how you could do it using dropna() and drop_duplicated() I have used my own sample data with one column with no values.

import pandas as pd
from io import StringIO

csv = StringIO('''2001,1,,a,a
        2001,2,,b,b
        2001,3,,c,c
        2005,1,,a,a
        2005,1,,c,c''')
df = pd.read_csv(csv, header=None )
print(df)

df will look like this

      0  1   2  3  4
0  2001  1 NaN  a  a
1  2001  2 NaN  b  b
2  2001  3 NaN  c  c
3  2005  1 NaN  a  a
4  2005  1 NaN  c  c

Then drop all columns ( how='all' ) where all values are na(NaN)

df_new = df.dropna(how='all', axis=1)

Take a transpose of the dataframe, the duplicate columns will become duplicate rows. Then use drop_duplicates on it to drop duplicate rows. Transpose it back to get your original data, without empty columns and duplicate columns.

df_new = df_new.T.drop_duplicates().T
df_new.columns = range(len(df_new.columns))
print(df_new)