inst/scripts/99.biomass_index_carstm_paper.r
In jae0/bio.snowcrab: Snow crab assessment suite using aegis* and associated projects and data streams.
# -------------------------------------------------
# Snow crab --- Areal unit modelling Hurdle / Delta model
# combination of three models via posterior simulation
# 1. Poisson on positive valued numbers offset by swept are
# 2. Meansize in space and time
# 3 Presence-absence
# the convolution of all three after simulation is called a Hurdle or Delta model
# -------------------------------------------------
# TODO::: move plotting calls to self-contained functions:
# -------------------------------------------------
# Part 1 -- construct basic parameter list defining the main characteristics of the study
source( file.path( code_root, "bio_startup.R" ) )
require(bio.snowcrab) # loadfunctions("bio.snowcrab")
year.assessment = 2023
yrs = 1999:year.assessment
spec_bio = bio.taxonomy::taxonomy.recode( from="spec", to="parsimonious", tolookup=2526 )
snowcrab_filter_class = "fb" # fishable biomass (including soft-shelled ) "m.mat" "f.mat" "imm"
# snowcrab_filter_class = "imm" # note poisson will not work due to var inflation .. nbinomial is a better choice
# snowcrab_filter_class = "f.mat"
# snowcrab_filter_class = "m.mat"
carstm_model_label= paste( "default", snowcrab_filter_class, sep="_" )
# params for number
pN = snowcrab_parameters(
project_class="carstm",
yrs=yrs,
areal_units_type="tesselation",
family = switch( snowcrab_filter_class,
imm = "nbinomial",
f.mat= "nbinomial",
m.mat= "nbinomial",
fb = "nbinomial",
"poisson"),
carstm_model_label= carstm_model_label,
selection = list(
type = "number",
biologicals=list( spec_bio=spec_bio ),
biologicals_using_snowcrab_filter_class=snowcrab_filter_class
)
)
# params for mean size .. mostly the same as pN
pW = snowcrab_parameters(
project_class="carstm",
yrs=yrs,
areal_units_type="tesselation",
family = "gaussian",
carstm_model_label= carstm_model_label,
selection = list(
type = "meansize",
biologicals=list( spec_bio=spec_bio ),
biologicals_using_snowcrab_filter_class=snowcrab_filter_class
)
)
# params for probability of observation
pH = snowcrab_parameters(
project_class="carstm",
yrs=yrs,
areal_units_type="tesselation",
family = "binomial", # "binomial", # "nbinomial", "betabinomial", "zeroinflatedbinomial0" , "zeroinflatednbinomial0"
carstm_model_label= carstm_model_label,
selection = list(
type = "presence_absence",
biologicals=list( spec_bio=spec_bio ),
biologicals_using_snowcrab_filter_class=snowcrab_filter_class
)
)
if (areal_units) {
# polygon structure:: create if not yet made
# for (au in c("cfanorth", "cfasouth", "cfa4x", "cfaall" )) plot(polygon_managementareas( species="snowcrab", au))
xydata = snowcrab.db( p=pN, DS="areal_units_input", redo=TRUE )
xydata = snowcrab.db( p=pN, DS="areal_units_input" )
sppoly = areal_units( p=pN, xydata=xydata[ which(xydata$yr %in% pN$yrs), ], redo=TRUE, verbose=TRUE ) # create constrained polygons with neighbourhood as an attribute
sppoly=areal_units( p=pN )
plot(sppoly["npts"])
sppoly$dummyvar = ""
xydata = st_as_sf( xydata, coords=c("lon","lat") )
st_crs(xydata) = st_crs( projection_proj4string("lonlat_wgs84") )
additional_features = snowcrab_mapping_features(pN) # for mapping below
tmap_mode("plot")
plt =
tm_shape(sppoly) +
tm_borders(col = "slategray", alpha = 0.5, lwd = 0.5) +
tm_shape( xydata ) + tm_sf() +
additional_features +
tm_compass(position = c("right", "TOP"), size = 1.5) +
tm_scale_bar(position = c("RIGHT", "BOTTOM"), width =0.1, text.size = 0.5) +
tm_layout(frame = FALSE, scale = 2) +
tm_shape( st_transform(polygons_rnaturalearth(), st_crs(sppoly) )) +
tm_borders(col = "slategray", alpha = 0.5, lwd = 0.5)
dev.new(width=14, height=8, pointsize=20)
plt
}
if (temperature_figures) {
# area-specific figures
# /home/jae/bio.data/bio.snowcrab/assessments/2022/timeseries/temperature_bottom.pdf
figure_area_based_extraction_from_carstm(DS="temperature", year.assessment=year.assessment ) # can only do done once we have an sppoly for snow crab
# full domain:
# default paramerters (copied from 03_temperature_carstm.R )
require(aegis.temperature)
params = list(
temperature = temperature_parameters(
project_class="carstm",
yrs=1970:year.assessment,
carstm_model_label="default"
)
)
sppoly=areal_units( p=pN )
pL = aegis.temperature::temperature_parameters( project_class="carstm", carstm_model_label="default" , yrs=p$yrs )
LUT= aegis_survey_lookuptable( aegis_project="temperature",
project_class="carstm", DS="carstm_predictions", pL=pL )
tss = aegis_lookup(
pl=pL, LUT=LUT,
LOCS=expand.grid( AUID=sppoly$AUID, timestamp= yrs + 0.75 ), LOCS_AU=sppoly,
project_class="carstm", output_format="areal_units",
variable_name=list( "predictions" ), statvars=c("mean", "sd"), space_resolution=pN$pres,
returntype = "data.table"
)
}
sppoly=areal_units( p=pN )
M = snowcrab.db( p=pN, DS="carstm_inputs", sppoly=sppoly, redo=TRUE ) # will redo if not found
additional_features = snowcrab_mapping_featuresres(pN) # for mapping below
# ------------------------------------------------
# Part 2 -- spatiotemporal statistical model
if ( spatiotemporal_model ) {
# total numbers
sppoly = areal_units( p=pN )
M = snowcrab.db( p=pN, DS="carstm_inputs", sppoly=sppoly ) # will redo if not found
io = which(M$tag=="observations")
ip = which(M$tag=="predictions")
iq = unique( c( which( M$totno > 0), ip ) )
iw = unique( c( which( M$totno > 5), ip ) ) # need a good sample to estimate mean size
# number
res = NULL; gc()
res = carstm_model( p=pN, data=M[ iq, ], sppoly=sppoly,
space_id = sppoly$AUID,
time_id = pN$yrs,
cyclic_id = pN$cyclic_levels,
# redo_fit=FALSE,
# debug = "summary",
theta=c(1.544, 2.744, 1.626, -2.066, 3.979, 0.683, 4.899, 4.395, 0.768, -5.015, 1.204, -2.771, 1.526),
nposteriors=5000,
posterior_simulations_to_retain=c( "summary", "random_spatial", "predictions"),
verbose=TRUE,
# redo_fit=FALSE,
# debug = "summary",
num.threads="4:3"
)
# mean size
res = NULL; gc()
res = carstm_model( p=pW, data=M[ iw, ], sppoly = sppoly,
space_id = sppoly$AUID,
time_id = pW$yrs,
cyclic_id = pW$cyclic_levels,
theta=c(6.309, 7.943, 1.805, 1.646, 8.713, 3.879, 13.561, 10.775, 6.306, -0.488, 6.004, -2.065, 1.391 ),
nposteriors=5000,
posterior_simulations_to_retain=c( "summary", "random_spatial", "predictions"),
verbose=TRUE,
num.threads="4:3",
# redo_fit=FALSE,
# debug = "summary",
control.inla = list( strategy="laplace", int.strategy="eb" )
)
# model pa using all data
res = NULL; gc()
res = carstm_model( p=pH, data=M, sppoly=sppoly,
space_id = sppoly$AUID,
time_id = pH$yrs,
cyclic_id = pH$cyclic_levels,
theta = c(1.007, 1.793, -4.280, 0.707, -3.044, 2.929, 3.479, -1.262, -1.965, -0.510, -2.144, 2.893 ),
nposteriors=5000,
posterior_simulations_to_retain=c( "summary", "random_spatial", "predictions"),
verbose=TRUE,
num.threads="4:3",
#redo_fit=FALSE,
# debug = "summary",
# control.family=list(control.link=list(model="logit")), # default for binomial .. no need to specify
control.inla = list( strategy="laplace", int.strategy="eb" )
)
# choose:
p = pN
p = pW
p = pH
if (0) {
# extract results
fit = carstm_model( p=p, DS="modelled_fit", sppoly = sppoly ) # extract currently saved model fit
fit$summary$dic$dic
fit$summary$dic$p.eff
plot(fit)
plot(fit, plot.prior=TRUE, plot.hyperparameters=TRUE, plot.fixed.effects=FALSE )
plot( fit, plot.prior=TRUE, plot.hyperparameters=TRUE, plot.fixed.effects=FALSE )
plot( fit$marginals.hyperpar$"Phi for space_time", type="l") # posterior distribution of phi nonspatial dominates
plot( fit$marginals.hyperpar$"Precision for space_time", type="l")
plot( fit$marginals.hyperpar$"Precision for setno", type="l")
fit = NULL
}
res = carstm_model( p=p, DS="carstm_modelled_summary", sppoly = sppoly ) # to load currently saved results
if (0) {
p = pH
res = carstm_model( p=p, DS="carstm_modelled_summary", sppoly = sppoly ) # to load currently saved results
o = carstm_2D_effects_probability(
res,
xvar = "inla.group(t, method = \"quantile\", n = 11)",
yvar = "inla.group(z, method = \"quantile\", n = 11)",
xgrid = seq( -1, 10.5, by=0.5),
ygrid = seq( 25, 350, by=25),
xslice = 4,
yslice = -200,
nx=200, ny=200,
theta = 140,
phi = 15
)
# use a larger domain than sppoly for the following estimate:
# sppoly is constrained to sampled locations, and so missing a lot of the inshore areas
x11()
crs_plot = st_crs( sppoly )
domain = polygon_managementareas( species="maritimes" )
domain = st_transform( domain, crs_plot )
data_mask = st_union( sppoly[which(sppoly$filter==1),1] )
# all = st_union( domain, data_mask )
nearshore = st_cast( st_difference( domain, data_mask ), "POLYGON")[1]
domain_new = st_union( data_mask, nearshore )
o = carstm_optimal_habitat(
res = res,
xvar = "inla.group(t, method = \"quantile\", n = 11)",
yvar = "inla.group(z, method = \"quantile\", n = 11)",
depths=switch( snowcrab_filter_class,
fb = c(100, 350),
imm = c( 160, 350),
f.mat = c(100, 160),
m.mat = c(160, 300)
),
probability_limit = 0.25,
nsims = 100,
domain=domain_new
)
dev.new();
print( o["depth_plot"] )
if (0) {
u = aegis::read_write_fast('/home/jae/tmp/temp_depth_habitat.RDS')
dev.new()
plot( habitat~yr, u, type="b", ylim=c(0.1, 0.33))
lines( habitat_lb~yr, u)
lines( habitat_ub~yr, u)
abline(v=1993)
abline(v=2012)
dev.new()
plot( habitat_sa~yr, u, type="b" )
lines( habitat_sa_lb~yr, u)
lines( habitat_sa_ub~yr, u)
abline(v=1993)
abline(v=2012)
ll = loess(habitat~yr, u, span=0.25 )
pp = predict( ll, u )
lines(pp ~ u$yr)
}
outputdir = file.path( p$modeldir, p$carstm_model_label )
fn_optimal = file.path( outputdir, "optimal_habitat_temperature_depth_effect.RDS" )
read_write_fast( data=o, file=fn_optimal )
o = aegis::read_write_fast(fn_optimal)
library(ggplot2)
dev.new(width=14, height=8, pointsize=20)
ggplot( o[["temperature_depth"]], aes(yr, habitat ) ) +
geom_ribbon(aes(ymin=habitat_lb, max=habitat_ub), alpha=0.2, colour=NA) +
geom_line() +
labs(x="Year", y="Habitat probabtility", size = rel(1.5)) +
# scale_y_continuous( limits=c(0, 300) )
theme_light( base_size = 22 )
dev.new(width=14, height=8, pointsize=20)
ggplot( o[["temperature_depth"]], aes(yr, habitat_sa ) ) +
geom_ribbon(aes(ymin=habitat_sa_lb, max=habitat_sa_ub), alpha=0.2, colour=NA) +
geom_line() +
labs(x="Year", y=bquote("Habitat surface area;" ~ km^2), size = rel(1.5)) +
# scale_y_continuous( limits=c(0, 300) )
theme_light( base_size = 22 )
}
if (0) {
# quick plots
vn=c( "random", "space", "re_total" )
vn=c( "random", "spacetime", "re_total" )
vn="predictions" # numerical density (km^-2)
tmatch= as.character(year.assessment)
carstm_map( res=res, vn=vn, tmatch=tmatch,
sppoly = sppoly,
palette="-RdYlBu",
plot_elements=c( "compass", "scale_bar", "legend" ),
additional_features=additional_features,
title =paste( vn, paste0(tmatch, collapse="-"), "no/m^2" )
)
# map all :
if ( number ) {
p=pN
outputdir = file.path( p$modeldir, p$carstm_model_label, "predicted.numerical.densities" )
ylab = "Number"
if ( !file.exists(outputdir)) dir.create( outputdir, recursive=TRUE, showWarnings=FALSE )
fn_root_prefix = "Predicted_numerical_abundance"
# title= paste( snowcrab_filter_class, "Number; no./m^2" )
}
if ( meansize) {
p=pW
ylab = "Mean weight"
outputdir = file.path( p$modeldir, p$carstm_model_label, "predicted.meansize" )
if ( !file.exists(outputdir)) dir.create( outputdir, recursive=TRUE, showWarnings=FALSE )
fn_root_prefix = "Predicted_meansize"
# title= paste( snowcrab_filter_class, "Mean weight; kg" )
}
if ( presence_absence ) {
p=pH
ylab = "Probability"
outputdir = file.path( p$modeldir, p$carstm_model_label, "predicted.presence_absence" )
if ( !file.exists(outputdir)) dir.create( outputdir, recursive=TRUE, showWarnings=FALSE )
fn_root_prefix = "Predicted_presence_absence"
title= paste( snowcrab_filter_class, "Probability")
}
if (0) {
# choose:
#p = pN
#p = pW
#p = pH
# extract results
fit = carstm_model( p=p, DS="modelled_fit", sppoly = sppoly ) # extract currently saved model fit
fit$summary$dic$dic
fit$summary$dic$p.eff
plot(fit)
plot(fit, plot.prior=TRUE, plot.hyperparameters=TRUE, plot.fixed.effects=FALSE )
plot( fit, plot.prior=TRUE, plot.hyperparameters=TRUE, plot.fixed.effects=FALSE )
plot( fit$marginals.hyperpar$"Phi for space_time", type="l") # posterior distribution of phi nonspatial dominates
plot( fit$marginals.hyperpar$"Precision for space_time", type="l")
plot( fit$marginals.hyperpar$"Precision for setno", type="l")
fit = carstm_model( p=pN, DS="modelled_fit")
fit = carstm_model( p=pH, DS="modelled_fit")
fit = carstm_model( p=pW, DS="modelled_fit")
res_vars = c( names( fit$hypers), names(fit$fixed) )
for (i in 1:length(res_vars) ) {
o = carstm_prior_posterior_compare( fit, vn=res_vars[i] )
dev.new(); print(o)
}
fit = NULL
}
oeffdir = file.path(p$modeldir, p$carstm_model_label, "figures")
carstm_plot_marginaleffects( p=p, outputdir=oeffdir, fn_root_prefix=fn_root_prefix )
additional_features = features_to_add(
p=p,
# area_lines="cfa.regions",
isobaths=c( 100, 200, 300, 400, 500 ),
xlim=c(-80,-40),
ylim=c(38, 60) # ,redo=TRUE
)
# map all bottom temps:
outputdir = file.path(p$modeldir, p$carstm_model_label, "maps" )
fn_root_prefix = "Substrate grainsize (mm)"
carstm_plot_map( p=p, outputdir=outputdir,
additional_features=additional_features,
toplot="random_spatial", probs=c(0.025, 0.975), transf=log10,
colors=rev(RColorBrewer::brewer.pal(5, "RdYlBu")) )
carstm_plot_map( p=p, outputdir=outputdir, additional_features=additional_features,
toplot="predictions", colors=rev(RColorBrewer::brewer.pal(5, "RdYlBu")) )
} #end each seletion
} # end spatiotemporal model
# ----------------------
# Part 3: assimilation of models
assimilate_numbers_and_size = TRUE
if (assimilate_numbers_and_size ) {
# wgts_max = 1.1 # kg, hard upper limit
# N_max = NULL
# # quantile( M$totno[ipositive]/M$data_offset[ipositive], probs=0.95, na.rm=TRUE )
# posterior sims
sims = carstm_posterior_simulations( pN=pN, pW=pW, pH=pH, pa_threshold=0.05, qmax=0.99 )
sims = sims / 10^6 # units: kg ; div (10^6) -> kt ;;
SM = aggregate_simulations(
sims=sims,
sppoly=sppoly,
fn=carstm_filenames( pN, returnvalue="filename", fn="aggregated_timeseries" ),
yrs=pN$yrs,
method="sum",
redo=TRUE
)
# units: kt/km^2
if (0) {
# to compute habitat prob
sims = carstm_posterior_simulations( pH=pH, pa_threshold=0.05, qmax=0.99 )
SM = aggregate_simulations(
sims=sims,
sppoly=sppoly,
fn=carstm_filenames( pN, returnvalue="filename", fn="aggregated_timeseries" ),
yrs=pN$yrs,
method="mean",
redo=TRUE
)
outputdir = file.path( carstm_filenames( pN, returnvalue="output_directory"), "aggregated_habitat_timeseries" )
RES= SM$RES
}
RES= SM$RES # units: kt
# RES = aggregate_simulations( fn=carstm_filenames( pN, returnvalue="filename", fn="aggregated_timeseries" ) )$RES
# note: using pN, even though this is biomass
outputdir = file.path( carstm_filenames( pN, returnvalue="output_directory"), "aggregated_biomass_timeseries" )
if ( !file.exists(outputdir)) dir.create( outputdir, recursive=TRUE, showWarnings=FALSE )
( fn = file.path( outputdir, "cfa_all.png") )
png( filename=fn, width=3072, height=2304, pointsize=12, res=300 )
plot( cfaall ~ yrs, data=RES, lty="solid", lwd=4, pch=20, col="slateblue", type="b", ylab="Biomass index (kt)", xlab="")
lines( cfaall_lb ~ yrs, data=RES, lty="dotted", lwd=2, col="slategray" )
lines( cfaall_ub ~ yrs, data=RES, lty="dotted", lwd=2, col="slategray" )
dev.off()
( fn = file.path( outputdir, "cfa_south.png") )
png( filename=fn, width=3072, height=2304, pointsize=12, res=300 )
plot( cfasouth ~ yrs, data=RES, lty="solid", lwd=4, pch=20, col="slateblue", type="b", ylab="Biomass index (kt)", xlab="")
lines( cfasouth_lb ~ yrs, data=RES, lty="dotted", lwd=2, col="slategray" )
lines( cfasouth_ub ~ yrs, data=RES, lty="dotted", lwd=2, col="slategray" )
dev.off()
( fn = file.path( outputdir, "cfa_north.png") )
png( filename=fn, width=3072, height=2304, pointsize=12, res=300 )
plot( cfanorth ~ yrs, data=RES, lty="solid", lwd=4, pch=20, col="slateblue", type="b", ylab="Biomass index (kt)", xlab="")
lines( cfanorth_lb ~ yrs, data=RES, lty="dotted", lwd=2, col="slategray" )
lines( cfanorth_ub ~ yrs, data=RES, lty="dotted", lwd=2, col="slategray" )
dev.off()
( fn = file.path( outputdir, "cfa_4x.png") )
png( filename=fn, width=3072, height=2304, pointsize=12, res=300 )
plot( cfa4x ~ yrs, data=RES, lty="solid", lwd=4, pch=20, col="slateblue", type="b", ylab="Biomass index (kt)", xlab="")
lines( cfa4x_lb ~ yrs, data=RES, lty="dotted", lwd=2, col="slategray" )
lines( cfa4x_ub ~ yrs, data=RES, lty="dotted", lwd=2, col="slategray" )
dev.off()
regions = c("cfanorth", "cfasouth", "cfa4x" )
region_label = c("N-ENS", "S-ENS", "4X")
a= cbind( "cfanorth", RES[,c("yrs", "cfanorth", "cfanorth_lb", "cfanorth_ub")] )
b= cbind( "cfasouth", RES[,c("yrs", "cfasouth", "cfasouth_lb", "cfasouth_ub")] )
c= cbind( "cfa4x", RES[,c("yrs", "cfa4x", "cfa4x_lb", "cfa4x_ub")] )
names(a) = names(b) = names(c) = c("region", "year", "mean", "lb", "ub")
tdb = rbind(a, b, c)
tdb$region = factor(tdb$region, levels=regions, labels =region_label)
tdb = tdb[(which(!is.na(tdb$region))), ]
fn = file.path( outputdir, "biomass_M0.png" )
require(ggplot2)
library(ggbreak)
color_map = c("#E69F00", "#56B4E9", "#CC79A7" )
out = ggplot(tdb, aes(x=year, y=mean, fill=region, colour=region)) +
geom_line( alpha=0.9, linewidth=1.2 ) +
geom_point(aes(shape=region), size=3, alpha=0.7 ) +
geom_errorbar(aes(ymin=lb,ymax=ub), linewidth=0.8, alpha=0.8, width=0.3) +
labs(x="Year/Année", y="Biomass index (kt) / Indice de biomasse (kt)", size = rel(1.5)) +
scale_colour_manual(values=color_map) +
scale_fill_manual(values=color_map) +
scale_shape_manual(values = c(15, 17, 19)) +
theme_light( base_size = 22) +
theme( legend.position="inside", legend.position.inside=c(0.75, 0.9), legend.title=element_blank()) +
scale_y_break(c(14, 28), scales = 1)
# scale_y_continuous( limits=c(0, 300) )
ggsave(filename=fn, plot=out, device="png", width=12, height = 8)
# map it ..mean density
sppoly = areal_units( p=pN ) # to reload
vn = paste("biomass", "predicted", sep=".")
outputdir = file.path( carstm_filenames( pN, returnvalue="output_directory"), "predicted_biomass_densities" )
if ( !file.exists(outputdir)) dir.create( outputdir, recursive=TRUE, showWarnings=FALSE )
B = apply( sims, c(1,2), mean ) # sims units (kt);
B[ which(!is.finite(B)) ] = NA
# brks = pretty( log10( quantile( B[], probs=c(0.05, 0.95), na.rm=TRUE )* 10^6) )
sa = units::drop_units(sppoly$au_sa_km2)
brks = pretty( ( quantile( log(B * 10^6 / sa), probs=c(0.05, 0.95), na.rm=TRUE )) )
additional_features = snowcrab_mapping_features(pN) # for mapping below
for (i in 1:length(pN$yrs) ) {
y = as.character( pN$yrs[i] )
# u = log10( B[,y]* 10^6 ) ## Total kt->kg: log10( kg )
u = log( B[,y]* 10^6 / sa) # ;; density log10( kg /km^2 )
sppoly[,vn] = u
outfilename = file.path( outputdir , paste( "biomass", y, "png", sep=".") )
plt = carstm_map( sppoly=sppoly, vn=vn,
breaks=brks,
additional_features=additional_features,
title=y,
# title=paste( "log_10( Predicted biomass density; kg/km^2 )", y ),
palette="-RdYlBu",
plot_elements=c( "compass", "scale_bar", "legend" ),
outfilename=outfilename
)
plt
}
} # end assimilate size and numbers
# end
jae0/bio.snowcrab documentation built on Nov. 6, 2024, 10:10 p.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
# -------------------------------------------------
# Snow crab --- Areal unit modelling Hurdle / Delta model
# combination of three models via posterior simulation
# 1. Poisson on positive valued numbers offset by swept are
# 2. Meansize in space and time
# 3 Presence-absence
# the convolution of all three after simulation is called a Hurdle or Delta model
# -------------------------------------------------
# TODO::: move plotting calls to self-contained functions:
# -------------------------------------------------
# Part 1 -- construct basic parameter list defining the main characteristics of the study
source( file.path( code_root, "bio_startup.R" ) )
require(bio.snowcrab) # loadfunctions("bio.snowcrab")
year.assessment = 2023
yrs = 1999:year.assessment
spec_bio = bio.taxonomy::taxonomy.recode( from="spec", to="parsimonious", tolookup=2526 )
snowcrab_filter_class = "fb" # fishable biomass (including soft-shelled ) "m.mat" "f.mat" "imm"
# snowcrab_filter_class = "imm" # note poisson will not work due to var inflation .. nbinomial is a better choice
# snowcrab_filter_class = "f.mat"
# snowcrab_filter_class = "m.mat"
carstm_model_label= paste( "default", snowcrab_filter_class, sep="_" )
# params for number
pN = snowcrab_parameters(
project_class="carstm",
yrs=yrs,
areal_units_type="tesselation",
family = switch( snowcrab_filter_class,
imm = "nbinomial",
f.mat= "nbinomial",
m.mat= "nbinomial",
fb = "nbinomial",
"poisson"),
carstm_model_label= carstm_model_label,
selection = list(
type = "number",
biologicals=list( spec_bio=spec_bio ),
biologicals_using_snowcrab_filter_class=snowcrab_filter_class
)
)
# params for mean size .. mostly the same as pN
pW = snowcrab_parameters(
project_class="carstm",
yrs=yrs,
areal_units_type="tesselation",
family = "gaussian",
carstm_model_label= carstm_model_label,
selection = list(
type = "meansize",
biologicals=list( spec_bio=spec_bio ),
biologicals_using_snowcrab_filter_class=snowcrab_filter_class
)
)
# params for probability of observation
pH = snowcrab_parameters(
project_class="carstm",
yrs=yrs,
areal_units_type="tesselation",
family = "binomial", # "binomial", # "nbinomial", "betabinomial", "zeroinflatedbinomial0" , "zeroinflatednbinomial0"
carstm_model_label= carstm_model_label,
selection = list(
type = "presence_absence",
biologicals=list( spec_bio=spec_bio ),
biologicals_using_snowcrab_filter_class=snowcrab_filter_class
)
)
if (areal_units) {
# polygon structure:: create if not yet made
# for (au in c("cfanorth", "cfasouth", "cfa4x", "cfaall" )) plot(polygon_managementareas( species="snowcrab", au))
xydata = snowcrab.db( p=pN, DS="areal_units_input", redo=TRUE )
xydata = snowcrab.db( p=pN, DS="areal_units_input" )
sppoly = areal_units( p=pN, xydata=xydata[ which(xydata$yr %in% pN$yrs), ], redo=TRUE, verbose=TRUE ) # create constrained polygons with neighbourhood as an attribute
sppoly=areal_units( p=pN )
plot(sppoly["npts"])
sppoly$dummyvar = ""
xydata = st_as_sf( xydata, coords=c("lon","lat") )
st_crs(xydata) = st_crs( projection_proj4string("lonlat_wgs84") )
additional_features = snowcrab_mapping_features(pN) # for mapping below
tmap_mode("plot")
plt =
tm_shape(sppoly) +
tm_borders(col = "slategray", alpha = 0.5, lwd = 0.5) +
tm_shape( xydata ) + tm_sf() +
additional_features +
tm_compass(position = c("right", "TOP"), size = 1.5) +
tm_scale_bar(position = c("RIGHT", "BOTTOM"), width =0.1, text.size = 0.5) +
tm_layout(frame = FALSE, scale = 2) +
tm_shape( st_transform(polygons_rnaturalearth(), st_crs(sppoly) )) +
tm_borders(col = "slategray", alpha = 0.5, lwd = 0.5)
dev.new(width=14, height=8, pointsize=20)
plt
}
if (temperature_figures) {
# area-specific figures
# /home/jae/bio.data/bio.snowcrab/assessments/2022/timeseries/temperature_bottom.pdf
figure_area_based_extraction_from_carstm(DS="temperature", year.assessment=year.assessment ) # can only do done once we have an sppoly for snow crab
# full domain:
# default paramerters (copied from 03_temperature_carstm.R )
require(aegis.temperature)
params = list(
temperature = temperature_parameters(
project_class="carstm",
yrs=1970:year.assessment,
carstm_model_label="default"
)
)
sppoly=areal_units( p=pN )
pL = aegis.temperature::temperature_parameters( project_class="carstm", carstm_model_label="default" , yrs=p$yrs )
LUT= aegis_survey_lookuptable( aegis_project="temperature",
project_class="carstm", DS="carstm_predictions", pL=pL )
tss = aegis_lookup(
pl=pL, LUT=LUT,
LOCS=expand.grid( AUID=sppoly$AUID, timestamp= yrs + 0.75 ), LOCS_AU=sppoly,
project_class="carstm", output_format="areal_units",
variable_name=list( "predictions" ), statvars=c("mean", "sd"), space_resolution=pN$pres,
returntype = "data.table"
)
}
sppoly=areal_units( p=pN )
M = snowcrab.db( p=pN, DS="carstm_inputs", sppoly=sppoly, redo=TRUE ) # will redo if not found
additional_features = snowcrab_mapping_featuresres(pN) # for mapping below
# ------------------------------------------------
# Part 2 -- spatiotemporal statistical model
if ( spatiotemporal_model ) {
# total numbers
sppoly = areal_units( p=pN )
M = snowcrab.db( p=pN, DS="carstm_inputs", sppoly=sppoly ) # will redo if not found
io = which(M$tag=="observations")
ip = which(M$tag=="predictions")
iq = unique( c( which( M$totno > 0), ip ) )
iw = unique( c( which( M$totno > 5), ip ) ) # need a good sample to estimate mean size
# number
res = NULL; gc()
res = carstm_model( p=pN, data=M[ iq, ], sppoly=sppoly,
space_id = sppoly$AUID,
time_id = pN$yrs,
cyclic_id = pN$cyclic_levels,
# redo_fit=FALSE,
# debug = "summary",
theta=c(1.544, 2.744, 1.626, -2.066, 3.979, 0.683, 4.899, 4.395, 0.768, -5.015, 1.204, -2.771, 1.526),
nposteriors=5000,
posterior_simulations_to_retain=c( "summary", "random_spatial", "predictions"),
verbose=TRUE,
# redo_fit=FALSE,
# debug = "summary",
num.threads="4:3"
)
# mean size
res = NULL; gc()
res = carstm_model( p=pW, data=M[ iw, ], sppoly = sppoly,
space_id = sppoly$AUID,
time_id = pW$yrs,
cyclic_id = pW$cyclic_levels,
theta=c(6.309, 7.943, 1.805, 1.646, 8.713, 3.879, 13.561, 10.775, 6.306, -0.488, 6.004, -2.065, 1.391 ),
nposteriors=5000,
posterior_simulations_to_retain=c( "summary", "random_spatial", "predictions"),
verbose=TRUE,
num.threads="4:3",
# redo_fit=FALSE,
# debug = "summary",
control.inla = list( strategy="laplace", int.strategy="eb" )
)
# model pa using all data
res = NULL; gc()
res = carstm_model( p=pH, data=M, sppoly=sppoly,
space_id = sppoly$AUID,
time_id = pH$yrs,
cyclic_id = pH$cyclic_levels,
theta = c(1.007, 1.793, -4.280, 0.707, -3.044, 2.929, 3.479, -1.262, -1.965, -0.510, -2.144, 2.893 ),
nposteriors=5000,
posterior_simulations_to_retain=c( "summary", "random_spatial", "predictions"),
verbose=TRUE,
num.threads="4:3",
#redo_fit=FALSE,
# debug = "summary",
# control.family=list(control.link=list(model="logit")), # default for binomial .. no need to specify
control.inla = list( strategy="laplace", int.strategy="eb" )
)
# choose:
p = pN
p = pW
p = pH
if (0) {
# extract results
fit = carstm_model( p=p, DS="modelled_fit", sppoly = sppoly ) # extract currently saved model fit
fit$summary$dic$dic
fit$summary$dic$p.eff
plot(fit)
plot(fit, plot.prior=TRUE, plot.hyperparameters=TRUE, plot.fixed.effects=FALSE )
plot( fit, plot.prior=TRUE, plot.hyperparameters=TRUE, plot.fixed.effects=FALSE )
plot( fit$marginals.hyperpar$"Phi for space_time", type="l") # posterior distribution of phi nonspatial dominates
plot( fit$marginals.hyperpar$"Precision for space_time", type="l")
plot( fit$marginals.hyperpar$"Precision for setno", type="l")
fit = NULL
}
res = carstm_model( p=p, DS="carstm_modelled_summary", sppoly = sppoly ) # to load currently saved results
if (0) {
p = pH
res = carstm_model( p=p, DS="carstm_modelled_summary", sppoly = sppoly ) # to load currently saved results
o = carstm_2D_effects_probability(
res,
xvar = "inla.group(t, method = \"quantile\", n = 11)",
yvar = "inla.group(z, method = \"quantile\", n = 11)",
xgrid = seq( -1, 10.5, by=0.5),
ygrid = seq( 25, 350, by=25),
xslice = 4,
yslice = -200,
nx=200, ny=200,
theta = 140,
phi = 15
)
# use a larger domain than sppoly for the following estimate:
# sppoly is constrained to sampled locations, and so missing a lot of the inshore areas
x11()
crs_plot = st_crs( sppoly )
domain = polygon_managementareas( species="maritimes" )
domain = st_transform( domain, crs_plot )
data_mask = st_union( sppoly[which(sppoly$filter==1),1] )
# all = st_union( domain, data_mask )
nearshore = st_cast( st_difference( domain, data_mask ), "POLYGON")[1]
domain_new = st_union( data_mask, nearshore )
o = carstm_optimal_habitat(
res = res,
xvar = "inla.group(t, method = \"quantile\", n = 11)",
yvar = "inla.group(z, method = \"quantile\", n = 11)",
depths=switch( snowcrab_filter_class,
fb = c(100, 350),
imm = c( 160, 350),
f.mat = c(100, 160),
m.mat = c(160, 300)
),
probability_limit = 0.25,
nsims = 100,
domain=domain_new
)
dev.new();
print( o["depth_plot"] )
if (0) {
u = aegis::read_write_fast('/home/jae/tmp/temp_depth_habitat.RDS')
dev.new()
plot( habitat~yr, u, type="b", ylim=c(0.1, 0.33))
lines( habitat_lb~yr, u)
lines( habitat_ub~yr, u)
abline(v=1993)
abline(v=2012)
dev.new()
plot( habitat_sa~yr, u, type="b" )
lines( habitat_sa_lb~yr, u)
lines( habitat_sa_ub~yr, u)
abline(v=1993)
abline(v=2012)
ll = loess(habitat~yr, u, span=0.25 )
pp = predict( ll, u )
lines(pp ~ u$yr)
}
outputdir = file.path( p$modeldir, p$carstm_model_label )
fn_optimal = file.path( outputdir, "optimal_habitat_temperature_depth_effect.RDS" )
read_write_fast( data=o, file=fn_optimal )
o = aegis::read_write_fast(fn_optimal)
library(ggplot2)
dev.new(width=14, height=8, pointsize=20)
ggplot( o[["temperature_depth"]], aes(yr, habitat ) ) +
geom_ribbon(aes(ymin=habitat_lb, max=habitat_ub), alpha=0.2, colour=NA) +
geom_line() +
labs(x="Year", y="Habitat probabtility", size = rel(1.5)) +
# scale_y_continuous( limits=c(0, 300) )
theme_light( base_size = 22 )
dev.new(width=14, height=8, pointsize=20)
ggplot( o[["temperature_depth"]], aes(yr, habitat_sa ) ) +
geom_ribbon(aes(ymin=habitat_sa_lb, max=habitat_sa_ub), alpha=0.2, colour=NA) +
geom_line() +
labs(x="Year", y=bquote("Habitat surface area;" ~ km^2), size = rel(1.5)) +
# scale_y_continuous( limits=c(0, 300) )
theme_light( base_size = 22 )
}
if (0) {
# quick plots
vn=c( "random", "space", "re_total" )
vn=c( "random", "spacetime", "re_total" )
vn="predictions" # numerical density (km^-2)
tmatch= as.character(year.assessment)
carstm_map( res=res, vn=vn, tmatch=tmatch,
sppoly = sppoly,
palette="-RdYlBu",
plot_elements=c( "compass", "scale_bar", "legend" ),
additional_features=additional_features,
title =paste( vn, paste0(tmatch, collapse="-"), "no/m^2" )
)
# map all :
if ( number ) {
p=pN
outputdir = file.path( p$modeldir, p$carstm_model_label, "predicted.numerical.densities" )
ylab = "Number"
if ( !file.exists(outputdir)) dir.create( outputdir, recursive=TRUE, showWarnings=FALSE )
fn_root_prefix = "Predicted_numerical_abundance"
# title= paste( snowcrab_filter_class, "Number; no./m^2" )
}
if ( meansize) {
p=pW
ylab = "Mean weight"
outputdir = file.path( p$modeldir, p$carstm_model_label, "predicted.meansize" )
if ( !file.exists(outputdir)) dir.create( outputdir, recursive=TRUE, showWarnings=FALSE )
fn_root_prefix = "Predicted_meansize"
# title= paste( snowcrab_filter_class, "Mean weight; kg" )
}
if ( presence_absence ) {
p=pH
ylab = "Probability"
outputdir = file.path( p$modeldir, p$carstm_model_label, "predicted.presence_absence" )
if ( !file.exists(outputdir)) dir.create( outputdir, recursive=TRUE, showWarnings=FALSE )
fn_root_prefix = "Predicted_presence_absence"
title= paste( snowcrab_filter_class, "Probability")
}
if (0) {
# choose:
#p = pN
#p = pW
#p = pH
# extract results
fit = carstm_model( p=p, DS="modelled_fit", sppoly = sppoly ) # extract currently saved model fit
fit$summary$dic$dic
fit$summary$dic$p.eff
plot(fit)
plot(fit, plot.prior=TRUE, plot.hyperparameters=TRUE, plot.fixed.effects=FALSE )
plot( fit, plot.prior=TRUE, plot.hyperparameters=TRUE, plot.fixed.effects=FALSE )
plot( fit$marginals.hyperpar$"Phi for space_time", type="l") # posterior distribution of phi nonspatial dominates
plot( fit$marginals.hyperpar$"Precision for space_time", type="l")
plot( fit$marginals.hyperpar$"Precision for setno", type="l")
fit = carstm_model( p=pN, DS="modelled_fit")
fit = carstm_model( p=pH, DS="modelled_fit")
fit = carstm_model( p=pW, DS="modelled_fit")
res_vars = c( names( fit$hypers), names(fit$fixed) )
for (i in 1:length(res_vars) ) {
o = carstm_prior_posterior_compare( fit, vn=res_vars[i] )
dev.new(); print(o)
}
fit = NULL
}
oeffdir = file.path(p$modeldir, p$carstm_model_label, "figures")
carstm_plot_marginaleffects( p=p, outputdir=oeffdir, fn_root_prefix=fn_root_prefix )
additional_features = features_to_add(
p=p,
# area_lines="cfa.regions",
isobaths=c( 100, 200, 300, 400, 500 ),
xlim=c(-80,-40),
ylim=c(38, 60) # ,redo=TRUE
)
# map all bottom temps:
outputdir = file.path(p$modeldir, p$carstm_model_label, "maps" )
fn_root_prefix = "Substrate grainsize (mm)"
carstm_plot_map( p=p, outputdir=outputdir,
additional_features=additional_features,
toplot="random_spatial", probs=c(0.025, 0.975), transf=log10,
colors=rev(RColorBrewer::brewer.pal(5, "RdYlBu")) )
carstm_plot_map( p=p, outputdir=outputdir, additional_features=additional_features,
toplot="predictions", colors=rev(RColorBrewer::brewer.pal(5, "RdYlBu")) )
} #end each seletion
} # end spatiotemporal model
# ----------------------
# Part 3: assimilation of models
assimilate_numbers_and_size = TRUE
if (assimilate_numbers_and_size ) {
# wgts_max = 1.1 # kg, hard upper limit
# N_max = NULL
# # quantile( M$totno[ipositive]/M$data_offset[ipositive], probs=0.95, na.rm=TRUE )
# posterior sims
sims = carstm_posterior_simulations( pN=pN, pW=pW, pH=pH, pa_threshold=0.05, qmax=0.99 )
sims = sims / 10^6 # units: kg ; div (10^6) -> kt ;;
SM = aggregate_simulations(
sims=sims,
sppoly=sppoly,
fn=carstm_filenames( pN, returnvalue="filename", fn="aggregated_timeseries" ),
yrs=pN$yrs,
method="sum",
redo=TRUE
)
# units: kt/km^2
if (0) {
# to compute habitat prob
sims = carstm_posterior_simulations( pH=pH, pa_threshold=0.05, qmax=0.99 )
SM = aggregate_simulations(
sims=sims,
sppoly=sppoly,
fn=carstm_filenames( pN, returnvalue="filename", fn="aggregated_timeseries" ),
yrs=pN$yrs,
method="mean",
redo=TRUE
)
outputdir = file.path( carstm_filenames( pN, returnvalue="output_directory"), "aggregated_habitat_timeseries" )
RES= SM$RES
}
RES= SM$RES # units: kt
# RES = aggregate_simulations( fn=carstm_filenames( pN, returnvalue="filename", fn="aggregated_timeseries" ) )$RES
# note: using pN, even though this is biomass
outputdir = file.path( carstm_filenames( pN, returnvalue="output_directory"), "aggregated_biomass_timeseries" )
if ( !file.exists(outputdir)) dir.create( outputdir, recursive=TRUE, showWarnings=FALSE )
( fn = file.path( outputdir, "cfa_all.png") )
png( filename=fn, width=3072, height=2304, pointsize=12, res=300 )
plot( cfaall ~ yrs, data=RES, lty="solid", lwd=4, pch=20, col="slateblue", type="b", ylab="Biomass index (kt)", xlab="")
lines( cfaall_lb ~ yrs, data=RES, lty="dotted", lwd=2, col="slategray" )
lines( cfaall_ub ~ yrs, data=RES, lty="dotted", lwd=2, col="slategray" )
dev.off()
( fn = file.path( outputdir, "cfa_south.png") )
png( filename=fn, width=3072, height=2304, pointsize=12, res=300 )
plot( cfasouth ~ yrs, data=RES, lty="solid", lwd=4, pch=20, col="slateblue", type="b", ylab="Biomass index (kt)", xlab="")
lines( cfasouth_lb ~ yrs, data=RES, lty="dotted", lwd=2, col="slategray" )
lines( cfasouth_ub ~ yrs, data=RES, lty="dotted", lwd=2, col="slategray" )
dev.off()
( fn = file.path( outputdir, "cfa_north.png") )
png( filename=fn, width=3072, height=2304, pointsize=12, res=300 )
plot( cfanorth ~ yrs, data=RES, lty="solid", lwd=4, pch=20, col="slateblue", type="b", ylab="Biomass index (kt)", xlab="")
lines( cfanorth_lb ~ yrs, data=RES, lty="dotted", lwd=2, col="slategray" )
lines( cfanorth_ub ~ yrs, data=RES, lty="dotted", lwd=2, col="slategray" )
dev.off()
( fn = file.path( outputdir, "cfa_4x.png") )
png( filename=fn, width=3072, height=2304, pointsize=12, res=300 )
plot( cfa4x ~ yrs, data=RES, lty="solid", lwd=4, pch=20, col="slateblue", type="b", ylab="Biomass index (kt)", xlab="")
lines( cfa4x_lb ~ yrs, data=RES, lty="dotted", lwd=2, col="slategray" )
lines( cfa4x_ub ~ yrs, data=RES, lty="dotted", lwd=2, col="slategray" )
dev.off()
regions = c("cfanorth", "cfasouth", "cfa4x" )
region_label = c("N-ENS", "S-ENS", "4X")
a= cbind( "cfanorth", RES[,c("yrs", "cfanorth", "cfanorth_lb", "cfanorth_ub")] )
b= cbind( "cfasouth", RES[,c("yrs", "cfasouth", "cfasouth_lb", "cfasouth_ub")] )
c= cbind( "cfa4x", RES[,c("yrs", "cfa4x", "cfa4x_lb", "cfa4x_ub")] )
names(a) = names(b) = names(c) = c("region", "year", "mean", "lb", "ub")
tdb = rbind(a, b, c)
tdb$region = factor(tdb$region, levels=regions, labels =region_label)
tdb = tdb[(which(!is.na(tdb$region))), ]
fn = file.path( outputdir, "biomass_M0.png" )
require(ggplot2)
library(ggbreak)
color_map = c("#E69F00", "#56B4E9", "#CC79A7" )
out = ggplot(tdb, aes(x=year, y=mean, fill=region, colour=region)) +
geom_line( alpha=0.9, linewidth=1.2 ) +
geom_point(aes(shape=region), size=3, alpha=0.7 ) +
geom_errorbar(aes(ymin=lb,ymax=ub), linewidth=0.8, alpha=0.8, width=0.3) +
labs(x="Year/Année", y="Biomass index (kt) / Indice de biomasse (kt)", size = rel(1.5)) +
scale_colour_manual(values=color_map) +
scale_fill_manual(values=color_map) +
scale_shape_manual(values = c(15, 17, 19)) +
theme_light( base_size = 22) +
theme( legend.position="inside", legend.position.inside=c(0.75, 0.9), legend.title=element_blank()) +
scale_y_break(c(14, 28), scales = 1)
# scale_y_continuous( limits=c(0, 300) )
ggsave(filename=fn, plot=out, device="png", width=12, height = 8)
# map it ..mean density
sppoly = areal_units( p=pN ) # to reload
vn = paste("biomass", "predicted", sep=".")
outputdir = file.path( carstm_filenames( pN, returnvalue="output_directory"), "predicted_biomass_densities" )
if ( !file.exists(outputdir)) dir.create( outputdir, recursive=TRUE, showWarnings=FALSE )
B = apply( sims, c(1,2), mean ) # sims units (kt);
B[ which(!is.finite(B)) ] = NA
# brks = pretty( log10( quantile( B[], probs=c(0.05, 0.95), na.rm=TRUE )* 10^6) )
sa = units::drop_units(sppoly$au_sa_km2)
brks = pretty( ( quantile( log(B * 10^6 / sa), probs=c(0.05, 0.95), na.rm=TRUE )) )
additional_features = snowcrab_mapping_features(pN) # for mapping below
for (i in 1:length(pN$yrs) ) {
y = as.character( pN$yrs[i] )
# u = log10( B[,y]* 10^6 ) ## Total kt->kg: log10( kg )
u = log( B[,y]* 10^6 / sa) # ;; density log10( kg /km^2 )
sppoly[,vn] = u
outfilename = file.path( outputdir , paste( "biomass", y, "png", sep=".") )
plt = carstm_map( sppoly=sppoly, vn=vn,
breaks=brks,
additional_features=additional_features,
title=y,
# title=paste( "log_10( Predicted biomass density; kg/km^2 )", y ),
palette="-RdYlBu",
plot_elements=c( "compass", "scale_bar", "legend" ),
outfilename=outfilename
)
plt
}
} # end assimilate size and numbers
# end
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.