Function Mapping (pandas)

pandas

  • Series.apply(function)
    • function’s parameter : multiple individual values (first parameter should be Series’ element itself)
    • Elements from Series are calculated by the function
  • DataFrame.applymap(function)
    • function’s parameter : a single individual value
    • Elements from DataFrame are calculated by the function
  • DataFrame.apply(function)
    • function’s parameter : a Series
    • Each column (axis=0) or row (axis=1) is calculated by the function
  • DataFrame.pipe(function)
    • function’s parameter : a DataFrame
    • Dataframe itself is calculated by the function

Time Series (pandas)

pandas

  • Array elements’ dtypes : Timestamp or Period
    • Basic structure : year-month-day (ex. 2024-01-02)
    • Timestamp : At time point
      • form : y-m-d or y-m-d time
    • Period : time interval
      • form : y or y-m or y-m-d or y-m-d time
  • If we set time series (one column) as index of dataframe, it is convenient to deal with these
    • only extracting year series or month series or day series
    • accessing the specific records using time series elements

Normalization (pandas)

pandas

  • If data range of a specific column is so wide, it can have influence on the result of machine learning more than other columns.
  • We should make the data range from each column be 0 ~ 1 or -1 ~ 1

Pagination


© 2017. All rights reserved.

Powered by Hydejack v조현진