Using spark ts ARIMA models to impute missing values in time series data

80 views

Skip to first unread message

Mahdi Bazargani

unread,

Mar 27, 2018, 10:08:52 PM3/27/18

to Time Series for Spark (the spark-ts package)

Hi,

I have a time series data with missing values, I want to impute missing values using ARIMA model as implemented in spark-ts package

say

ts: 12 , NA, 13, 14 , 18,20

the package should fill the NA

I was used to use R impute-ts package particular kalman filter with auto-arima model but not sure if I can get similar results with different ARIMA models currently exist in Spark-ts?

Reply all

Reply to author

Forward

0 new messages