Iso cor

[NDurden]

Modifying and using calibrate_carbon() to calibrate NEON isotope data (see flow.isoCo2.cor.R for the workflow).

[rfiorella]

A few issues to resolve:

• [x] use of ::: calls cause note in R CMD check, and will cause CRAN to reject future versions. Should remove before merging.
• [x] Need to update r-lib GitHub actions scripts to v2 to fix GHA R-CMD-check and code coverage failures
• [x] Test if changes to R/calibrate_carbon.R are necessary - much of this additional behavior should be automatically handled by neonUtilities stackEddy

[rfiorella]

@NDurden I've gone through the changes to the package functions, and I think I do see the need now for the changes made to calibrate_carbon.R (stackEddy will take a folder of several files, or a single hdf5 file, but I don't think it works on a list of files, at least based on the documentation).

I think the right place for this code though is ingest_data function and not calibrate_carbon.R, to keep calibrate_carbon.R as short as possible.

The other possibility is that this change shouldn't be in NEONiso at all, but that neonUtilities::stackEddy could be altered in order to handle a list of hdf5 files. @cklunch do you think there would be other users that might need to load, for example, a year of eddy data that isn't in its own folder or in a zip file?

[cklunch]

@rfiorella stackEddy() does work on a vector of filepaths that each point to an H5 file. I haven't really advertised that option because you miss out on some of the metadata if you do it that way - e.g. the updates in v2.3.0 that give you the science review flags and the citation info won't work, because those metadata are sourced from other files in the download package. Although of course that is also true if you stack from a single H5 file, I guess in that case it seems like less of a big deal since people presumably know they're working with minimal data.

Try it out, let me know if it works for your use case!

[rfiorella]

@cklunch Sure thing, I'll write a couple tests and see if it works as expected for this case. Thanks!

@NDurden I'll be merging this PR if you're satisfied with the changes I've made after adding these tests and moving the code to ingest data. Let me know if you have any concerns!

[NDurden]

@rfiorella sounds good to me and let me know if any changes that I need to make. Thanks!

[NDurden]

@rfiorella thanks for working on these. I will test my workflow and will let you know how that goes.

[NDurden]

@rfiorella I tested my workflow with latest merge and it gave me the error messages 'separate_wider_delim' is not an exported object from 'namespace:tidyr'. The error seems to occur in stackEddy. I tested on neonUtilities (version 2.3.0). Do you have any idea what happened?

[rfiorella]

@NDurden hmm, I have not seen that error message before. What version of tidyr is installed on the machine you're working on? I think these functions are quite new in tidyr, so you might try updating tidyr if it is < 1.3.0

@cklunch any other ideas?

[NDurden]

@rfiorella @cklunch that error definitely came from the old version of tidyr. However, it gave me another error message which came from the site name (https://github.com/NEONScience/NEON-utilities/blob/f1473397404e1a596d6b69f9d8c16ee740b8372a/neonUtilities/R/stackEddy.R#L450) I will try to work around by adjusting my workflow. I will keep you posted.

[cklunch]

@NDurden Make sure all the packages neonUtilities is dependent on are up to date, if they're not it can definitely cause problems. If you still run into issues let me know. Thanks!