reprex VS workflowr

Provide us with the code in a minimum working example (MWE) form (typically called reprex in the R community). If the process of producing this doesn't help you solve the issue, it at least allows us to (a) copy and paste to run the code, and (b) far more likely to be able to spot the error(s) than when they’re embedded in a load of code that isn’t necessary to highlight the problem. These should be placed in a code block, either every line indented 4 spaces or within backticks. (Credit to Mooks79 for this.) More info here: https://www.reddit.com/r/learnpython/wiki/faq#wiki_how_do_i_format_code.3F

Let me start by saying it's generally preferred that you offer a reprex in a post like this, meaning at a minimum you should share (1) the minimal necessary code to reproduce your issue and (2) at least a toy data set that can be used to test your code (so, not a screen shot but maybe something created by dput(WSD), dpasta(WSD), or the like). I personally would prefer that you also format your code as a code block in your post to make it easier to read, but that's not a deal breaker (just me being fastidious).

Is reproducible example would be a nice start: https://reprex.tidyverse.org/

It's really hard to help you debug without a reprex

df_sl = data.frame( SubLvlList = c( '1-Z', '2-Z', '1-Y', '2-Y' ) ) data.frame( ReturnMatch = sapply( df_tl[,1], function(x) { df_sl[,1][sapply(df_sl[,1], function(y) {grepl(y, x)})] }, USE.NAMES = FALSE) ) #> ReturnMatch #> 1 1-Z #> 2 1-Z #> 3 2-Z #> 4 2-Z #> 5 1-Y #> 6 1-Y #> 7 2-Y #> 8 2-Y #> 9 1-Z #> 10 1-Z #> 11 2-Z #> 12 2-Z #> 13 1-Y #> 14 1-Y #> 15 2-Y #> 16 2-Y ``` Created on 2023-03-25 with [reprex v2.0.2](https://reprex.tidyverse.org)

Making a reproducible example is a super good way to check that you are running what you think you are running.

reprex::reprex()

What you have posted is not valid R code. See https://reprex.tidyverse.org/ for example on how to create a reproducible example.

``` Created on 2022-11-11 with reprex v2.0.2

https://reprex.tidyverse.org A sample of the code might help in understanding what happened

I would start by saying "bioinformatics world" is a very broad term. Most of it does not involve managing multiple models, which is what MLflow seems to be for. Most work is cleaning the data and interpreting the results, which is highly project-specific. Something like workflowr is generally more appropriate, but even that is an overkill for most people.

Targets stores all your data serialised on disk, and only loads them in as needed by the dependencies. For a standardised folder structure, you can take inspiration from workflowr.