frbcesab / popbayes

:package: Bayesian Model to Estimate Animal Population Trends from Counts Series

popbayes

The goal of the R package popbayes is to infer trends of one or several populations over time from series of counts. It does so by accounting for count precision (provided or inferred based on expert knowledge, e.g. guesstimates), smoothing the population rate of increase over time, and accounting for the maximum demographic potential of species. Inference is carried out in a Bayesian framework. This work is part of the FRB-CESAB working group AfroBioDrivers.

Installation

Before using this package, users need to install the freeware JAGS.

You can install the stable version of the package from the CRAN with:

install.packages("popbayes")

Alternatively you can install the development version from GitHub with:

# install.packages("remotes")

remotes::install_github("frbcesab/popbayes", build_vignettes = TRUE)

Overview

Get started

Please read the Get started vignette.

Citation

Please cite this package as:

Casajus N. & Pradel R. (2023) popbayes: Bayesian model to estimate population trends from counts series. R package version 1.2.0. URL: https://frbcesab.github.io/popbayes/.

You can also run:

citation("popbayes")

## A BibTeX entry for LaTeX users is:
## 
## @Manual{,
##   title  = {{popbayes}: {B}ayesian model to estimate population trends from counts series,
##   author = {{Casajus N.}, and {Pradel R.}},
##   year   = {2023},
##   note   = {R package version 1.2.0},
##   url    = {https://frbcesab.github.io/popbayes/},
## }

popbayes's Issues

Rename conversion_fact in conversion_A2G

Rename conversion_fact in conversion_A2G

Rename year in date

Rename year in date

Default value for path in list_series()

Add a default value for path in list_series()

Implement weight_to_rmax function

Implement weight_to_rmax function

Use count instead of counts

Use count instead of counts (including diagram)

Return error if lower_ci == upper_ci

Return error if lower_ci == upper_ci

make fields pref_field_method, field_method, conversion_A2G optional in format_data

currently, they are assigned a default value. This apparently prevents them from being absent in case for instance of homogeneous series.

Write vignette

Release popbayes 1.1

First release:

• usethis::use_cran_comments()
• Update (aspirational) install instructions in README
• Proofread Title: and Description:
• Check that all exported functions have @return and @examples
• Check that Authors@R: includes a copyright holder (role 'cph')
• Check licensing of included files
• Review https://github.com/DavisVaughan/extrachecks

Prepare for release:

• devtools::build_readme()
• urlchecker::url_check()
• devtools::check(remote = TRUE, manual = TRUE)
• devtools::check_win_devel()
• rhub::check_for_cran()
• Review pkgdown reference index for, e.g., missing topics

Submit to CRAN:

• usethis::use_version('major')
• devtools::submit_cran()
• Approve email

Wait for CRAN...

• Accepted 🎉
• usethis::use_github_release()
• usethis::use_dev_version()
• usethis::use_news_md()
• Add CRAN badge in README
• Add installation instructions in README
• Tweet

years order?

It should be checked if years must be sorted or not.
First year receives a specific treatment in model.

problems with Tibbles

in format_data, if argument data is a tibble, it will pass the test "data must be a data.frame", but will fail the test
"The column date must be numeric" (subsetting rules are different and yield a tibble instead of a vector.

Suggested solution: after the initial test "data must be a data.frame", add the line
data <- as.data.frame(data)

priority of information for field method, preferred field method, conversion_A2G, and rmax

currently, info has precedence; followed by the imported data, and eventually the internal package information. As flexibility is desirable at the series level (the same species in different environmental may be applied different values for these fields), precedence should be given to the imported data as this is the only level where such distinction is possible.
This also means changing format_data() to check consistency at the series level rather than the species level

Allow different CI coverage in data

Allow different CI coverage in data.

Allow NA and duplicates in data$conversion_fact...

Just check that there is only one unique non NA value.
Same for data$pref_field_method, data$rmax..

Release popbayes 1.0

First release:

• usethis::use_cran_comments()
• Update (aspirational) install instructions in README
• Proofread Title: and Description:
• Check that all exported functions have @return and @examples
• Check that Authors@R: includes a copyright holder (role 'cph')
• Check licensing of included files
• Review https://github.com/DavisVaughan/extrachecks

Prepare for release:

• devtools::build_readme()
• urlchecker::url_check()
• devtools::check(remote = TRUE, manual = TRUE)
• devtools::check_win_devel()
• rhub::check_for_cran()
• Review pkgdown reference index for, e.g., missing topics

Submit to CRAN:

• usethis::use_version('major')
• devtools::submit_cran()
• Approve email

Wait for CRAN...

• Accepted 🎉
• usethis::use_github_release()
• usethis::use_dev_version()
• usethis::use_news_md()
• Add CRAN badge in README
• Add installation instructions in README
• Tweet

role of info and category to clarify

Is info used to carry the info about some species provided with the package?
If so, may it be modified by the user?
If not, info is provided by the user and should take precedence on info in the package itself.
info is just a more convenient way to bring information that is situated at the species level and would otherwise appear on several rows in a count series?

In case the user can provide only some info on some species, this should be possible.

A column "category" should be allowed for a new species for which there is no more detailed info.

Implement code testings

Edit documentation

field method should be allowed to be NA for guestimates

Extract rmax only in format_data

Extract rmax only in format_data()

Control error for info dataset in format_data()

Control error for info dataset in format_data():

• Missing species in info
• Missing column in info
• Allow additional column(s) in info.

Convert Garamba dataset to CSV

Convert Garamba dataset to CSV to explain users how to import their own data.

X rather than G for Guesstimate

'G' is used as the code for guesstimates in stat_method. But 'G' is also used as the code for 'ground' among the field methods. This leads to confusions. Replace 'G' by 'X' as a reminder for 'eXpert knowledge' in the stat methods.

problems with Tibbles

in format_data, if argument data is a tibble, it will pass the test "data must be a data.frame", but will fail the test
"The column date must be numeric" (subsetting rules are different and yield a tibble instead of a vector).

Suggested solution: after the initial test "data must be a data.frame", add the line
data <- as.data.frame(data)

saving several models for the same data set

Currently only the last model for a given data set is kept. Sometimes, one may want to try different options. these may be different degrees of smoothing, the use or not of rmax. Playing with rmax values. Playing with the conversion factor... Results are saved in a subdirectory built from the species and area names, meaning that the last model fitted overwrite the previous one. An enhancement would be to allow saving all models. This can already be done in R by saving the outputs from fit_trend in different objects but these objects cannot be manipulated easily.

A first improvement would be to allow output of fit_trend as input for plot_trend.

Create unique species_info data

Create unique species_info data merging conversion_data and rmax_data

Add Zenodo + covr

Need to find an easy way to pass Rmax values

With all information in one dataset, need to find an easy way to pass Rmax values (old structure does not fit now).

Release popbayes 1.2.0

Prepare for release:

• git pull
• Check current CRAN check results
• Polish NEWS
• usethis::use_github_links()
• urlchecker::url_check()
• devtools::build_readme()
• devtools::check(remote = TRUE, manual = TRUE)
• devtools::check_win_devel()
• revdepcheck::revdep_check(num_workers = 4)
• Update cran-comments.md
• git push
• Draft blog post

Submit to CRAN:

• usethis::use_version('minor')
• devtools::submit_cran()
• Approve email

Wait for CRAN...

• Accepted 🎉
• Add preemptive link to blog post in pkgdown news menu
• usethis::use_github_release()
• usethis::use_dev_version(push = TRUE)
• Finish blog post
• Tweet

not unique years

These can be fractional (e.g. 2001.66) and duplicated (2 counts in the same year is allowed). Yet, the output of format_data mentions "The number of unique years".
Need to check that there is no problem with this.