tt_test() errs with relatively high number of habitats
maurolepore opened this issue · 1 comments
maurolepore commented
library(tidyverse)
#> Warning: package 'purrr' was built under R version 3.5.3
library(fgeo)
#> -- Attaching packages -------------------------------------------- fgeo 1.1.2.9000 --
#> v fgeo.analyze 1.1.4 v fgeo.tool 1.2.2
#> v fgeo.plot 1.1.3.9000 v fgeo.x 1.1.2
#> -- Conflicts ---------------------------------------------------- fgeo_conflicts() --
#> x fgeo.tool::filter() masks dplyr::filter(), stats::filter()
#> x dplyr::intersect() masks base::intersect()
#> x dplyr::lag() masks stats::lag()
#> x dplyr::setdiff() masks base::setdiff()
#> x dplyr::setequal() masks base::setequal()
#> x dplyr::union() masks base::union()
set.seed(1)
# Small dataset from Luquillo
cns_luq <- fgeo.x::tree6_3species
hab_luq <- fgeo.x::habitat
sp_top3_luq <- unique(fgeo.x::tree6_3species$sp)
pdim_luq <- c(320, 500)
gsize_luq <- 20
# Passes
hab <- fgeo_habitat(fgeo.x::elevation, 20, 8)
as_tibble(tt_test(cns_luq, hab))
#> Using `plotdim = c(320, 500)`. To change this value see `?tt_test()`.
#> Using `gridsize = 20`. To change this value see `?tt_test()`.
#> # A tibble: 24 x 8
#> habitat sp N.Hab Gr.Hab Ls.Hab Eq.Hab Rep.Agg.Neut Obs.Quantile
#> * <chr> <chr> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
#> 1 1 CASARB 11 587 1012 1 0 0.367
#> 2 2 CASARB 16 1433 163 4 0 0.896
#> 3 3 CASARB 2 1200 368 32 0 0.75
#> 4 4 CASARB 14 886 710 4 0 0.554
#> 5 5 CASARB 6 223 1365 12 0 0.139
#> 6 6 CASARB 6 418 1177 5 0 0.261
#> 7 7 CASARB 5 932 658 10 0 0.582
#> 8 8 CASARB 6 1423 171 6 0 0.889
#> 9 1 PREMON 49 1223 374 3 0 0.764
#> 10 2 PREMON 37 627 968 5 0 0.392
#> # ... with 14 more rows
autoplot(hab)
# Errs
hab <- fgeo_habitat(fgeo.x::elevation, 20, 16)
as_tibble(tt_test(cns_luq, hab))
#> Using `plotdim = c(320, 500)`. To change this value see `?tt_test()`.
#> Using `gridsize = 20`. To change this value see `?tt_test()`.
#> Error in if (spprophab[i] > Torspprophab[i]) {: missing value where TRUE/FALSE needed
autoplot(hab)
Created on 2019-03-22 by the reprex package (v0.2.1)
maurolepore commented
Relates to #40 (comment)
The issue is not a given empty quadrat, but empty habitats (set of quadrats) that perfectly match with a habitat of one of the translated maps
library(tidyverse)
library(fgeo)
#> -- Attaching packages ------------------------------------------- fgeo 1.1.2 --
#> v fgeo.analyze 1.1.4.9000 v fgeo.tool 1.2.2
#> v fgeo.plot 1.1.3.9000 v fgeo.x 1.1.2
#> -- Conflicts ---------------------------------------------- fgeo_conflicts() --
#> x fgeo.tool::filter() masks dplyr::filter(), stats::filter()
#> x dplyr::intersect() masks base::intersect()
#> x dplyr::lag() masks stats::lag()
#> x dplyr::setdiff() masks base::setdiff()
#> x dplyr::setequal() masks base::setequal()
#> x dplyr::union() masks base::union()
cns_luq <- fgeo.x::tree6_3species
hab <- fgeo_habitat(fgeo.x::elevation, 20, 16)
as_tibble(tt_test(cns_luq, hab))
#> Using `plotdim = c(320, 500)`. To change this value see `?tt_test()`.
#> Using `gridsize = 20`. To change this value see `?tt_test()`.
#> Every quadrat in the dataset must have at lease one tree.
#> For more information see https://github.com/forestgeo/fgeo.analyze/issues/40.
Created on 2019-04-26 by the reprex package (v0.2.1)