plantphys / spectratrait

A tutorial R package for illustrating how to fit, evaluate, and report spectra-trait PLSR models. The package provides functions to enhance the base functionality of the R pls package, identify an optimal number of PLSR components, standardize model validation, and vignette examples that utilize datasets sourced from EcoSIS (ecosis.org)
GNU General Public License v3.0
12 stars 9 forks source link

Additional pre-submission updates #38

Closed serbinsh closed 4 years ago

serbinsh commented 4 years ago

Updated R_Scripts/functions.R to revise names of component selection options. Updated R_Scripts/expanded_spectra-trait_reseco_lma_plsr_example.R to allow for selection of custom output. Added additional output plotting.

serbinsh commented 4 years ago

@neo0351 sorry making more changes I will let you know when ready. Adding a fix for something @JulienLamour just found

serbinsh commented 4 years ago

@neo0351 note updated/bug fixed component selection. its pretty fast now because we are no longer running internal cross validation during the selection (duh) and instead using the fraction left off of when subseting the calibration data. So now I think we can really speed up the NEON run./

neo0351 commented 4 years ago

Great!!! Are you going to merge this soon? Or should I test it first? (is there a way to do that?)

serbinsh commented 4 years ago

@neo0351 merged. looking for testing now. FYI - lots of changes in expanded_spectra-trait_reseco_lma_plsr_example.R that created outputs to save for the user that I/we will need to incorporate into the rest of the scripts.

neo0351 commented 4 years ago

Let's test the one first, then we can edit the rest once we are happy.

serbinsh commented 4 years ago

Indeed. I also just realized we need to be sure we are saving coefficients properly in the new pls_permutation function. specifically that we have the intercepts and that they are the first part of each column? My brain is fried right now