Error in trainMatch Step: rawPredictionCol Vectors Must Have Length 2 but Got 1

·Jul 09, 2024 10:47 PM

During the trainMatch step I get this error - zingg.common.client.ZinggClientException: Exception thrown in awaitResult: requirement failed: rawPredictionCol vectors must have length=2, but got 1

3 comments

· Sorted by Oldest

Help Zingg

Error in trainMatch Step: rawPredictionCol Vectors Must Have Length 2 but Got 1

Tim D.

·Jul 09, 2024 10:47 PM

During the trainMatch step I get this error - zingg.common.client.ZinggClientException: Exception thrown in awaitResult: requirement failed: rawPredictionCol vectors must have length=2, but got 1

3 comments

· Sorted by Oldest

Vikas G.
·
how many records did you label? how many match and how many not match?
Tim D.
·
The end goal is to only match some 300-400 unmatched 'new' entities (and primarily only 1-2 columns of data like Name + State) to a 'mastered' list of 12.5k. But there are nearly 400k of variations that already match to 12.5k entities. Can you guide me on how to structure this data and process for zingg to tackle this use case?
Vikas G.
·
did you run the findTrainingData phase and label phase

Vikas G.
·
how many records did you label? how many match and how many not match?
Tim D.
·
The end goal is to only match some 300-400 unmatched 'new' entities (and primarily only 1-2 columns of data like Name + State) to a 'mastered' list of 12.5k. But there are nearly 400k of variations that already match to 12.5k entities. Can you guide me on how to structure this data and process for zingg to tackle this use case?
Vikas G.
·
did you run the findTrainingData phase and label phase