gdrplatform / gDRcore

R package to process dose-response curve data with the GR methods
https://gdrplatform.github.io/gDRcore
1 stars 1 forks source link

`validate_mappings` errors inside of `create_SE` with default nested confounders #14

Closed ChristopherEeles closed 1 year ago

ChristopherEeles commented 1 year ago

Reprex (inside the Docker images):

library(gDR)

# Define path for data stored in gDR package
dataDir <- system.file("extdata", "data1", package = "gDRimport")

# Extract path for example raw_data
manifest <- list.files(dataDir, pattern = "manifest", full.names = TRUE)
template <- list.files(dataDir, pattern = "Template", full.names = TRUE)
raw_data <- list.files(dataDir, pattern = "^RawData", full.names = TRUE)

# Import data
imported_data <- import_data(manifest, template, raw_data)

se <- create_SE(imported_data)
## INFO [2022-08-16 20:44:51] 
## INFO [2022-08-16 20:44:51] 
## Error in { : 
##  task 1 failed - "error in evaluating the argument 'x' in selecting a method for function ## 'unique': subscript out of bounds"
## In addition: There were 24 warnings (use warnings() to see them)

The cause is an attempt to subset a data.frame or DataFrame via [[ with a length > 1 character vector on line 177 of https://github.com/gdrplatform/gDRcore/blob/master/gDRcore/R/create_SE.R.

I will work around by only using one nested confounder for now.

Best, Chris

bczech commented 1 year ago

Hi @ChristopherEeles, the bug was fixed in #15 .

Best, Bartek

ChristopherEeles commented 1 year ago

Hi @bczech,

Is there a pipeline setup to propagate bug fixes to the Docker images or will I need to reinstall manually?

Best, Chris

bczech commented 1 year ago

Hi @ChristopherEeles,

The best way is to just pull the latest Docker image from master and work on the up-to-date image.

Best, Bartek