FRESCALO implemented in R by Dr Jon Yearsley (original link here).
FRESCALO is a method to correct biological recording data for variation in recorder effort. The method is described in Hill, MO (2012) Local frequency as a key to interpreting species occurrence data when recording effort is not known, Methods in Ecology and Evolution, doi: 10.1111/j.2041-210X.2011.00146.x
The original method is a FORTRAN program. This R script takes the same input files as FRESCALO:
- a weights file (defining a neighbourhood around a focal region)
- a species file giving the species observed at each region, and the date of the observation
More information about FRESCALO at http://www.brc.ac.uk/biblio/frescalo-computer-program-analyse-your-biological-records and https://github.com/BiologicalRecordsCentre/sparta/wiki/frescalo
- Added calculation of standard deviations and fixed some bugs encountered with sparse data.
- Minor changes so that species that are absent from a time period are not dropped from the results.
- Added s_it and w (site x time period recording effort metrics) to the
trend()function outputs. - Added function to calculate species x sites x time period predicted occupancies (P_ijt).
- Included dataset and weights from sparta used to check all outputs. Matched against original fortran and this independent R coding based on direct translation from fortran: https://github.com/sacrevert/fRescalo
unicorn_TF.rdaincludes the results from running the original fortran code on the test dataclusterTestDat.csvwith the weights inGB_LC_Wts.txt.
Under development, but see https://rpubs.com/sacrevert/parallelFrescalo (code in this repo here). TO DO:
- Bayesian meta-regression example