TheAnachronism/docspell
1
0
Fork 0
mirror of https://github.com/TheAnachronism/docspell.git synced 2025-05-31 20:42:51 +00:00
Code Issues Packages Projects Releases Wiki Activity

Don't use test data if there is just one config

Browse Source 
If classifier models cannot be compared, there is no reason to test.
This commit is contained in:
Eike Kettner 2020-09-07 20:01:30 +02:00
parent 2f9f981742
commit c9bd57592b
1 changed files with 2 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
modules/analysis/src/main/scala/docspell/analysis/nlp/StanfordTextClassifier.scala
View File

@ -66,7 +66,8 @@ final class StanfordTextClassifier[F[_]: Sync: ContextShift](
} yield res
def splitData(logger: Logger[F], in: RawData): F[TrainData] = {
val nTest = (in.count * 0.15).toLong
val f = if (cfg.classifierConfigs.size > 1) 0.15 else 0.0
val nTest = (in.count * f).toLong
val td =
TrainData(in.file.resolveSibling("train.txt"), in.file.resolveSibling("test.txt"))