search

XanderHorn / autoML

Automated machine learning in R
MIT License
9 stars 7 forks source link

just tried running your example code and I get an error #3

Closed lvsund closed 5 years ago

lvsund commented 5 years ago

code that was run:

library(autoEDA) library(autoML) library(data.table) library(caret)

set.seed(1991) full <- fread("/Volumes/NASPC/Downloads/kickstarter-projects/ks-projects-201801.csv", data.table = F, stringsAsFactors = F)

Create binary target

full <- subset(full, full$state %in% c("failed","successful"))

Sample down for decrease in run time ~ 99k obs

full <- full[caret::createDataPartition(y = full$state, p = 0.3, list = F),]

Quick tabular EDA

eda <- exploreData(x = full)

res <- autoML(train = full, target = "state", id = "ID", trainMode = "reduced", models = c("xgboost","randomforest"), modelInterpretability = F)

I get the following error:

autoPreProcess | Checking data autoPreProcess | Production code generated Error in [.data.frame(train, , target) : undefined columns selected In addition: Warning message: In if (id != "auto" & length(id) > 0) { : the condition has length > 1 and only the first element will be used

when i run against my own data I consistently get the Error in [.data.frame(train, , target) : undefined columns selected error

XanderHorn commented 5 years ago

This has been fixed, it was caused by a bug in the code. Thank you for reporting it, please download and re-install the package to run it again.

Best Xander