0

Is it possible to perform oneHotDecoder after using OneHotEncoder in spark ml? Is there any way to achieve this?

StringIndexer dateIndexer = new StringIndexer();
csvData = dateIndexer.setInputCol("Date").setOutputCol("dateIndex").fit(csvData).transform(csvData);
StringIndexer timeIndexer = new StringIndexer();
csvData = timeIndexer.setInputCol("Time").setOutputCol("timeIndex").fit(csvData).transform(csvData);
OneHotEncoderEstimator encoder = new OneHotEncoderEstimator(); csvData = encoder.setInputCols(new String[] { "dateIndex", "timeIndex"}).setOutputCols(new String[] { "dateVector", "timeVector"}).fit(csvData).transform(csvData);

I could not find any solution on this in the spark API docs.

zcoop98
  • 2,590
  • 1
  • 18
  • 31
Sudeep Nanda
  • 1
  • 2
  • 1
    Please provide a code sample of what you have attempted as a starting point for you to receive help – TyBourque Jul 27 '21 at 15:33
  • StringIndexer dateIndexer = new StringIndexer(); csvData=dateIndexer.setInputCol("Date") .setOutputCol("dateIndex") .fit(csvData) .transform(csvData); StringIndexer timeIndexer = new StringIndexer(); csvData=timeIndexer.setInputCol("Time") .setOutputCol("timeIndex") .fit(csvData) .transform(csvData);OneHotEncoderEstimator encoder = new OneHotEncoderEstimator(); csvData = encoder.setInputCols(new String[] {"dateIndex","timeIndex"}) .setOutputCols(new String[] {"dateVector","timeVector"}) .fit(csvData).transform(csvData); – Sudeep Nanda Jul 28 '21 at 10:39

0 Answers0