There's a lot of changed lines here but the reality is that the vast majority of the edits are moving existing code from one file to another.
I went through _targets.R and moved individual targets to more appropriate locations in (I think exclusively) 3_harmonize.R. I also emptied out the src/functions.R file and created new script files in 3_harmonize/src for those functions. Basically the overall goal here was to put the targets and their functions in locations corresponding to the part of the workflow they're used in. Some targets are left in _targets.R. These should be related to configuring the pipeline.
Additional changes:
I added subfolders (4_rs_inventory, 5_rs_download) for new, future steps in the AquaSat workflow
I added tar_cue("never") in the download step of 2_download.R in order to avoid new downloads of the WQP data unintentionally. I also added a step (wqp_data_aoi_formatted_filtered_out) that exports the WQP dataset as a feather file so that it can be used without running the whole download process
The SDD report rendering step took forever and was using up a lot of memory/erroring out sometimes so I changed the workflow to export SDD-related figures being used in the current report placeholder as PNG files instead of ggplot objects inside an RDS. This saves space/time when reading in the components needed in the report while still keeping it linked up into the logic of the pipeline
There's a lot of changed lines here but the reality is that the vast majority of the edits are moving existing code from one file to another.
I went through _targets.R and moved individual targets to more appropriate locations in (I think exclusively) 3_harmonize.R. I also emptied out the src/functions.R file and created new script files in 3_harmonize/src for those functions. Basically the overall goal here was to put the targets and their functions in locations corresponding to the part of the workflow they're used in. Some targets are left in _targets.R. These should be related to configuring the pipeline.
Additional changes:
tar_cue("never")
in the download step of 2_download.R in order to avoid new downloads of the WQP data unintentionally. I also added a step (wqp_data_aoi_formatted_filtered_out
) that exports the WQP dataset as a feather file so that it can be used without running the whole download process