Pandas

Applications in pandas

“pandas is kind of excel in python”

1. Dataframe

   # 创建一个示例 DataFrame
   df = pd.DataFrame({
   'A': [1, 2, 3],
   'B': [4, 5, 6]
   })

2. Dataframe to numpy:
   df.to_numpy()
3. Numpy to Dataframe:
   df = pd.Dataframe(array)
4. drop some columns:

   # drop by column name list
   column_list = ['1', '2', '3']
   df = df.drop(column_list, axis=1) # drop some columns

5. rename column:

   # rename by column name:
   df.rename(columns={"A":"a", "B":"b"})
   
   # rename by index
   df.rename(index={0: "x", 1: "y"})

6. get column name list:

   column_list = df.columns # return "Index['column1', 'column2'...'column10']"

7. get continuous part of dataframe

   df_part1 = df.iloc[:, :6] # first 5 columns of df
   
   df_part2 = df.iloc[:5, :] # first 5 rows of df
   
   df_index = df.iloc[1, 2] # element at index (1, 2)

8. modify a column

   df['column_name'] = df['column_name'].map({'ClassA': 1, 'ClassB': 2}) # mapping Label to Number
   
   df['column_name'] = df['column_name'].map(lambda x: x * 2) # mapping element using lambda

9. form a dataframe by multiple arrays

   df = pd.DataFrame({
      'Column1': array1,
      'Column2': array2
   })