leaves
leaves copied to clipboard
pure Go implementation of prediction part for GBRT (Gradient Boosting Regression Trees) models from popular frameworks
for Nan or given missing value of 0.0f, how to correctly deal with it by xgboost of go version
github.com/dmitryikh/leaves v0.0.0-20230708180554-25d19a787328 predict 57.46251861844212 python xgboost=0.90 predict 10.252338 this predict is correct python xgboost=1.7.2 predict 57.462517 my xgb trained by xgboost=0.90 (leaves load xgboost will get error when version is...
panic: runtime error: makeslice: len out of range goroutine 1 [running]: github.com/dmitryikh/leaves/internal/xgbin.ReadString(0x140000940c0) go/pkg/mod/github.com/dmitryikh/[email protected]/internal/xgbin/xgbin_io.go:176 +0x68 github.com/dmitryikh/leaves/internal/xgbin.ReadModelHeader(0x140000940c0) go/pkg/mod/github.com/dmitryikh/[email protected]/internal/xgbin/xgbin_io.go:230 +0x6c github.com/dmitryikh/leaves.XGEnsembleFromReader(0x140000940c0, 0x1) go/pkg/mod/github.com/dmitryikh/[email protected]/xgensemble_io.go:140 +0x40 github.com/dmitryikh/leaves.XGEnsembleFromFile({0x1026945ce?, 0x102605838?}, 0x1) go/pkg/mod/github.com/dmitryikh/[email protected]/xgensemble_io.go:259 +0x138 main.main() Documents/code/GoAI/main.go:98 +0x40 exit...
Currently `only version=v3 is supported` v4 was released a while ago https://github.com/microsoft/LightGBM/releases/tag/v4.0.0 Latest version is v4.4 https://github.com/microsoft/LightGBM/releases/tag/v4.4.0