knitr::opts_chunk$set(echo = TRUE)
This is a documenation of the cleaning round in October 2022 - summer 2023. Mainly serves to create a searchable commit. But also to keep track of changes.
TODO HERE go on with plotting things : currently : plot_modelwise_GDM
and plot_unique_GDM
. Need to create input for average over threshold plots, and either move to separate script or include in plot_modelwise_GDM.
TODO HERE GDM_multifun_thresholds : almost finished runningt he plotUncertainty. Need 2 more, and TODO TODO TODO important **GDM_multifun_thresholds.Rmd** : check if there is an error in the plotting function of gdm. Why do x values vary??
merge scripts alpha_plots.Rmd and introductory_plots.Rmd
[x] DONE address package error of a map-related package
plot_GDM.Rmd
:
removed automaticrun
variable : all runs are automatic from now.
chosing model from model_names
.
3 functions for the same job, just different dataset : create_bio_aboveground_barplot
, create_bio_belowground_barplot
, create_abio_barplot
. Lession learned for future project : do a common function for the same types of plots
reading in LUI input : before I gave the variable compon_lui
by hand, now this is called lui
and taken from model_names_selecton$lui
.
renamed to plot_modelwise_GDM. Only contains plots which are produced for each model indivdually. All other plots (that only need to be runned once = unique run) are in plot_unique_GDM
small datasets
small datasets that are stored in /vignettes are transferred to /data
usethis::use_data(\<nameofdataset\>)
if necessary with overwrite = T
to overwrite existing dataset
No loading is needed, because R packages use "lazy-loading", i.e. only load data when necessary.
model_names
will be used uniquely to select the current model to run, or to run all of them.
names need to be cleaned!
model_names column "name" renamed to "modelname"
added columns : funs, lui containing more info about the selected model --> reduce the amount of hand-filled input
model_names
column luiinput
will not be used in the future any more : there will not be a model with all 3 components together, but only with individual ones. (because LUI components are correlated.)
Dec 22 Implemented new calculation of average over all threshold : NEW is that the avearage is not only calcluated for the maximum values, but across all 200 values of each predictor. Therefore, the function plotUncertainty_sd
has been deprecated, and a new function plotUncertainty_slim
was implemented.
prepare_and_run_GDM.Rmd
implement the use of the model_names
table. Select used models with model_names_selection
.
funs
needs to be changed to model_names_selection$funs
compon_luli is deprecated: use lui
instead
replaced modelname
with model_names_selection$modelname
the_analysis_hive.Rmd
now. (hive-drone wording is used here)helper_datasets
from vignettes/
from data_assembly/helper_data and analysis/helper_data
usefulplotids
added to /data
colorder_gdm_input
added to /data
Evaluate if useful content is in there, think about what to do later.
Take care that the following points are included in re-analysis :
%var expl. plot is re-done. Seems there was an error in manually entering the values in a large table. New : use the small tables extracted for each model and rbind them to a full table of model specs.
Implement permutations in plot_GDM.Rmd
.
document data
screen through all scripts and solve remaining #TODO
remove sections_to-be_loaded
in the end of every script!
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.