Proposal of some updates to read_habitatmap_xxx() #140
Conversation
This makes it conform 2020, which seems to use a more logical ordering.
Co-authored-by: Cécile Herr <31855012+cecileherr@users.noreply.github.com>
Co-authored-by: Cécile Herr <31855012+cecileherr@users.noreply.github.com>
From https://github.com/inbo/n2khab-preprocessing/blob/b088c44/src/generate_habitatmap_stdized/10_generate_habmap_stdized.Rmd: > An exception to this rule are following codes: `3130,rbbmr`, `3140,rbbmr`, `3150,rbbmr` and `3160,rbbmr`. Also in the R code of above file, it can be seen it is only these codes that are handled.
R/read_habitatdata.R
Outdated
| #' created for each of them. | ||
| #' The variable \code{certain} in this case will be \code{TRUE} for both types. |
There was a problem hiding this comment.
About issue #117:
you may want to add a note on the interpretation of phab for code combinations from one column of the original habitatmap
Should also be mentioned in the geospatial hab vignette, but I think this can be useful here too, so:
| #' created for each of them. | |
| #' The variable \code{certain} in this case will be \code{TRUE} for both types. | |
| #' created for each of them, with \code{phab} for each new row simply | |
| #' set to the original value of \code{phab}. | |
| #' The variable \code{certain} in this case will be \code{TRUE} for both types. | |
| #' \itemize{ | |
| #' \item Note that this implies that the a given polygon could contain the same | |
| #' type with variable \code{certain} as \code{TRUE} several times, e.g. when: | |
| #' \code{31xx_rbbmr} is present with \code{phab} = yy% and | |
| #' \code{31xx} is present with \code{phab} = zz%. | |
| #' In that case the rows with the same \code{polygon_id}, \code{type} | |
| #' and \code{certain} were gathered into one row and the respective | |
| #' \code{phab} were summed up |
There was a problem hiding this comment.
Perfect, thanks! I'll re-add your suggestion as a new commit, as the lines it depends on are flagged as outdated (so cannot apply suggestion).
There was a problem hiding this comment.
I'm supposing this bullet is at a lower level. The {} of \itemize needed to be closed 😉
There was a problem hiding this comment.
@cecileherr Are you sure this is restricted to 31xx_rbbmr and it does not occur in general?
Anyway, the code executes this as a general rule. So I think it will be better to state it in general and add this bullet as a subnote under the first bullet (where you added 'with phab for each new row simply set to the original value of phab').
I leave consideration to you; I'll postpone the initial plan of adding the above suggestion.
There was a problem hiding this comment.
Sidenote: it appears that '%' needs to be escaped as \%
There was a problem hiding this comment.
@cecileherr Are you sure this is restricted to
31xx_rbbmrand it does not occur in general?Anyway, the code executes this as a general rule. So I think it will be better to state it in general and add this bullet as a subnote under the first bullet (where you added 'with phab for each new row simply set to the original value of phab').
I thought it was a general rule in the code, but that it only happened for the 31xx_rbbmr case, but I was wrong (e.g. 605723_v2014 is an example for the uncertain case). I would be tempted to add this is as 3rd item (after the uncertain case and the 31xx_rbbmr case), instead of making a subitem out of it.
Something like (other items abbreviated):
#' \itemize{
#' \item For some polygons the vegetation type is uncertain (...)
#' if only one vegetation type is provided.
#' \item Some polygons contain both a standing water habitat type (...)
#' created for each of them, with \code{phab} for each new row simply
#' set to the original value of \code{phab}.
#' The variable \code{certain} in this case will be \code{TRUE} for both types.
#' \item After those first two steps, a given polygon could contain the same
#' type with the same value for \code{certain} several times, e.g. when:
#' \code{31xx_rbbmr} is present with \code{phab} = yy\% and
#' \code{31xx} is present with \code{phab} = zz\%.
#' In that case the rows with the same \code{polygon_id}, \code{type}
#' and \code{certain} were gathered into one row and the respective
#' \code{phab} were summed up.
#' \item For some polygons the original vegetation code in the (...)
#' code was adjusted.
#' }
@florisvdh : your thought on this?
There was a problem hiding this comment.
Good idea @cecileherr. It better fits the idea of 'steps'.
Minor comments:
when:drop colonsummed up: means 'enumerated'; should become something like 'added up'.with the same value for \code{certain} several times: I think '[...] repeated several times' is better
I suggest you add another suggestion in this PR, or else do it as commit in your branch update_habitat2020_ch (including the Rd file update) after merging this PR. Merging this branch is up to you anyway, since its target branch is yours.
There was a problem hiding this comment.
Many thanks for the update (and sorry, I had obviously missed issue 117)
Maybe just take a look at this suggestion:
#140 (comment)
and then I think we will have covered all open questions, so we should be able to merge
|
Thank you for reviewing @cecileherr 🙏. |
Co-authored-by: Cécile Herr <31855012+cecileherr@users.noreply.github.com>
|
@cecileherr for me this PR is finished, you can merge |
Many thanks ! Let's go! |
filter_habto the front: in case of a mismatch the error will come up immediately. Otherwise thehabitatmapdata source is read first, which can take long.read_habitatmap_stdized(): adding the exception for"3130,rbbmr"etc. sincehabitatmap_stdized_2020_v1.Also addressed some remaining parts of issue #117. Specifically I propose for the 2018 versions to reorder the
xxx_typestibble columns for stdized/terr according to the order of the 2020 datasource version, where this was done in a more logical way. Do you agree? - Also pinging @ToonHub for this one.Commit 9eff6ec on documenting the key goes back to inbo/n2khab-preprocessing#50 (comment):
Results of testing the functions
Created on 2021-05-11 by the reprex package (v2.0.0)
Session info