Conversion in Column (pandas)

30 Dec 2023 in Study / Computer science on Pandas

pandas

• Conversion of unit
• Conversion of data type

Unit Conversion

• We make the Unit Conversion Factors
  • ex) kg → g : 1000 g / 1 kg
• original data * (Unit Conversion Factor) → data with new unit
• df[new column’s name] = df[original column’s name] * (Unit Conversion Factor)
  • add a new column from an original column to the dataframe

Data Type Conversion

• df[column’s name] = df[column’s name].astype(‘new dtypes’)
• There are lots of cases requiring the conversion of data type
  • Example case 1 : When the numerical data in a specific column which should be processed as number are stored as object because of some non-numerical data in the same column
    • 1^st : remove the rows having non-numerical data
    • 2^nd : df[column’s name] = df[column’s name].astype(‘float’) (or ‘int’)
    • If we wanna be back to object, df[column’s name] = df[column’s name].astype(‘str’)
  • Example case 2 : Category dtypes
    • wanna make the groups (Group name is the data in a specific column)
    • df[column’s name] = df[column’s name].astype(“category”)

Extra Knowledge

• round
  • df[column’s name] = df[column’s name].round(n) : At n location, it gonna happen
  • If n == 2, 1.879 → 1.9
• df.dtypes
  • return Series (index : names of columns, value : dtypes)
• df[column’s name].unique()
  • return list for the unique values of a specific column
• df[column’s name].replace({original data1 : new data1, original data2 : new data2, ….})
  • In a specific column, all of specific data (keys) are changed to “new data” (There gonna be automatically coversion of dtypes)
• df[column’s name].sample(n)
  • return Series that has n data selected from the dataframe