Changing Column (pandas)

27 Jan 2024 in Study / Computer science on Pandas

pandas

• Changing order
• Split

Changing Order

• Extract current column list
  • columns = list(df.columns.values)
• Sorting the column list
  • columns = sorted(list(df.columns.values))
    • Dictionary order
  • columns = list(reversed(list(df.columns.values)))
    • reverse order of orginal column list (not reverse order of dictionary)
• Create column list directly
  • columns = [‘c1’, ‘c2’, …]
• Apply the change of column list to dataframe
  • df = df[columns]

Split

• We wanna split one column into several columns (One datum of the one column should be separated, ex. 2001-05-20 → [2001, 05, 20])
• First of all, the data type should be ‘str’
  • df[‘c1’] = df[‘c1’].astype(‘str’)
• Second, we make Series representing the result list like [2001, 05, 20]
  • s = df[‘c1’].str.split(“special character”)
  • Based on special character, we split the data
• Third, we make Series which gonna be new column
  • df[‘new column’] = s.str.get(index)
  • ex. [year, month, days] → get(0) : year, get(1) : month, get(2) : days

import pandas as pd

df = pd.DataFrame({'c1' : ["1/28/30", "5/14/21", "10/20/30"], 'c2' : [40, 50, 50], 'c3' : [70, 50, 60]})

df['c1'] = df['c1'].astype('str')

s = df['c1'].str.split("/")

df['first'] = s.str.get(0)
df['second'] = s.str.get(1)
df['third'] = s.str.get(2)

print(s)
print()
print(df)

0     [1, 28, 30]
1     [5, 14, 21]
2    [10, 20, 30]
Name: c1, dtype: object

         c1  c2  c3 first second third
0   1/28/30  40  70     1     28    30
1   5/14/21  50  50     5     14    21
2  10/20/30  50  60    10     20    30