In this vignette we show how to use the do.fit = FALSE
flag to debug a WHAM model.
Load the ASAP3 data file.
wham.dir <- find.package("wham")
asap3 <- read_asap3_dat(file.path(wham.dir,"extdata","ex7_SNEMAYT.dat"))
Prepare the WHAM model (m3
from example
1).
input <- prepare_wham_input(asap3, recruit_model=2, model_name="Ex 1: SNEMA Yellowtail Flounder",
NAA_re = list(sigma="rec+1", cor="iid"))
Try to fit the model… uh oh.
mod <- fit_wham(input, do.osa = F, do.retro = F)
#> Order of parameters:
#> [1] "log_catch_sig_scale" "log_index_sig_scale" "log_N1_pars"
#> [4] "log_NAA_sigma" "trans_NAA_rho" "log_NAA"
#> [7] "logR_proj" "mean_rec_pars" "logit_q"
#> [10] "q_re" "q_repars" "q_prior_re"
#> [13] "logit_selpars" "selpars_re" "sel_repars"
#> [16] "catch_paa_pars" "index_paa_pars" "F_devs"
#> [19] "log_F1" "M_a" "M_re"
#> [22] "M_repars" "log_b" "Ecov_re"
#> [25] "Ecov_beta" "Ecov_process_pars" "Ecov_obs_logsigma"
#> [28] "Ecov_obs_sigma_par" "Ecov_obs_logsigma_re"
#> Not matching template order:
#> [1] "mean_rec_pars" "logit_q" "q_prior_re"
#> [4] "q_re" "q_repars" "log_F1"
#> [7] "F_devs" "log_N1_pars" "log_NAA_sigma"
#> [10] "trans_NAA_rho" "log_NAA" "logR_proj"
#> [13] "logit_selpars" "selpars_re" "sel_repars"
#> [16] "catch_paa_pars" "index_paa_pars" "M_a"
#> [19] "M_re" "M_repars" "log_b"
#> [22] "log_catch_sig_scale" "log_index_sig_scale" "Ecov_re"
#> [25] "Ecov_beta" "Ecov_process_pars" "Ecov_obs_logsigma"
#> [28] "Ecov_obs_logsigma_re" "Ecov_obs_sigma_par"
#> Your parameter list has been re-ordered.
#> (Disable this warning with checkParameterOrder=FALSE)
#> iter: 1 Error in iterate(par) :
#> Newton dropout because inner gradient had non-finite components.
#> Warning in stats::nlminb(model$par, model$fn, model$gr, control = list(iter.max
#> = 1000, : NA/NaN function evaluation
#> iter: 1 Error in iterate(par) :
#> Newton dropout because inner gradient had non-finite components.
#> Error in ff(x, order = 1) :
#> inner newton optimization failed during gradient calculation
#> outer mgc: NaN
#> Error in stats::nlminb(model$par, model$fn, model$gr, control = list(iter.max = 1000, : NA/NaN gradient evaluation
What’s wrong? It looks like the likelihood function is returning
NaN
. It is often easier to diagnose problems like this
using the unoptimized model, i.e. look at everything using the
initial parameter values. WHAM includes a do.fit = F
flag
in fit_wham
to return the unoptimized model object returned
by TMB::MakeADFun
. Let’s see how it works.
mod <- fit_wham(input, do.fit = F)
#> Order of parameters:
#> [1] "log_catch_sig_scale" "log_index_sig_scale" "log_N1_pars"
#> [4] "log_NAA_sigma" "trans_NAA_rho" "log_NAA"
#> [7] "logR_proj" "mean_rec_pars" "logit_q"
#> [10] "q_re" "q_repars" "q_prior_re"
#> [13] "logit_selpars" "selpars_re" "sel_repars"
#> [16] "catch_paa_pars" "index_paa_pars" "F_devs"
#> [19] "log_F1" "M_a" "M_re"
#> [22] "M_repars" "log_b" "Ecov_re"
#> [25] "Ecov_beta" "Ecov_process_pars" "Ecov_obs_logsigma"
#> [28] "Ecov_obs_sigma_par" "Ecov_obs_logsigma_re"
#> Not matching template order:
#> [1] "mean_rec_pars" "logit_q" "q_prior_re"
#> [4] "q_re" "q_repars" "log_F1"
#> [7] "F_devs" "log_N1_pars" "log_NAA_sigma"
#> [10] "trans_NAA_rho" "log_NAA" "logR_proj"
#> [13] "logit_selpars" "selpars_re" "sel_repars"
#> [16] "catch_paa_pars" "index_paa_pars" "M_a"
#> [19] "M_re" "M_repars" "log_b"
#> [22] "log_catch_sig_scale" "log_index_sig_scale" "Ecov_re"
#> [25] "Ecov_beta" "Ecov_process_pars" "Ecov_obs_logsigma"
#> [28] "Ecov_obs_logsigma_re" "Ecov_obs_sigma_par"
#> Your parameter list has been re-ordered.
#> (Disable this warning with checkParameterOrder=FALSE)
This runs without an error. The optimization failed because the
likelihood was NaN
, and now we can see which of the
likelihood components was responsible. To do so, we need to look at the
REPORT
ed objects with "nll"
in their name. We
can use the $report()
function from TMB.
therep = mod$report()
See all of the REPORT
ed objects.
names(therep)
#> [1] "pred_log_catch" "q_prior_re" "selAA"
#> [4] "sigma_q" "FAA_tot" "log_SPR0"
#> [7] "Ecov_beta" "log_FXSPR" "Ecov_obs_sigma_par"
#> [10] "NAA_devs" "mean_rec_pars" "QAA"
#> [13] "nll_catch_acomp" "M_a" "pred_catch_paa"
#> [16] "MAA" "log_Y_FXSPR_static" "SSB"
#> [19] "F" "nll_agg_indices" "nll"
#> [22] "log_YPR_FXSPR" "log_R_FXSPR_static" "nll_Ecov_obs"
#> [25] "selpars" "nll_Ecov_obs_sig" "Ecov_re"
#> [28] "log_SPR_FXSPR" "log_SPR0_static" "pred_indices"
#> [31] "nll_NAA" "log_FXSPR_static" "pred_IAA"
#> [34] "logit_q_mat" "pred_CAA" "nll_agg_catch"
#> [37] "nll_sel" "log_FXSPR_iter" "nll_q"
#> [40] "logit_selpars" "Fbar" "nll_Ecov"
#> [43] "log_FXSPR_iter_static" "Ecov_obs_sigma" "rho_q"
#> [46] "q_re" "Ecov_process_pars" "sel_repars"
#> [49] "nll_index_acomp" "ZAA" "pred_log_indices"
#> [52] "M_re" "selpars_re" "Ecov_x"
#> [55] "NAA" "pred_catch" "R_XSPR"
#> [58] "log_SSB_FXSPR" "M_repars" "log_SSB_FXSPR_static"
#> [61] "pred_NAA" "FAA" "log_SPR_FXSPR_static"
#> [64] "nll_q_prior" "log_Y_FXSPR" "pred_index_paa"
#> [67] "Ecov_out" "nll_M" "log_YPR_FXSPR_static"
#> [70] "q"
Now just get the objects with "nll"
in their name, and
sum over all individual values.
sapply(grep("nll",names(therep),value=T), function(x) sum(therep[[x]]))
#> nll_catch_acomp nll_agg_indices nll nll_Ecov_obs
#> 2844.2380 1562.8441 NaN 0.0000
#> nll_Ecov_obs_sig nll_NAA nll_agg_catch nll_sel
#> 0.0000 346.2991 NaN 0.0000
#> nll_q nll_Ecov nll_index_acomp nll_q_prior
#> 0.0000 0.0000 3921.8316 0.0000
#> nll_M
#> 0.0000
The likelihood components that are equal to 0 are not used in the
model (no random effects on M
, Ecov
,
selectivity
, etc.). nll
is the total
likelihood and is NaN
. The one troublesome component is
nll_agg_catch
. This is the total (aggregate) catch from the
fishery in each year. It could be an issue with the catch data or the
model-predicted catch, since both are in the likelihood calculation.
Search the WHAM
.cpp file for “nll_agg_catch”. We find that on line 888, this
depends on agg_catch
(the catch data) and
pred_catch
(model-predicted catch).
The catch data looks ok.
input$data$agg_catch
#> [,1]
#> [1,] 14549.0000
#> [2,] 17088.0000
#> [3,] 5732.0000
#> [4,] 3436.0000
#> [5,] 5223.0000
#> [6,] 8085.0000
#> [7,] 9883.0000
#> [8,] 8021.0000
#> [9,] 6607.0000
#> [10,] 15764.0000
#> [11,] 22211.0000
#> [12,] 11225.0000
#> [13,] 4817.0000
#> [14,] 4620.0000
#> [15,] 2652.0000
#> [16,] 2782.0000
#> [17,] 8349.0000
#> [18,] 17916.0000
#> [19,] 6430.0000
#> [20,] 2695.0000
#> [21,] 771.0008
#> [22,] 735.0004
#> [23,] 343.0004
#> [24,] 759.0001
#> [25,] 1222.0000
#> [26,] 1087.0000
#> [27,] 1403.0000
#> [28,] 1397.0000
#> [29,] 1449.0000
#> [30,] 945.0005
#> [31,] 666.0005
#> [32,] 619.0002
#> [33,] 346.0002
#> [34,] 396.0000
#> [35,] 502.0004
#> [36,] 583.0006
#> [37,] 453.0006
#> [38,] 290.9995
#> [39,] 390.0002
#> [40,] 563.0002
#> [41,] 645.9999
#> [42,] 625.0000
#> [43,] 337.0005
#> [44,] 151.9999
The predicted catch is calculated on lines 879-880.
pred_catch
depends on NAA
(numbers-at-age),
FAA
(F at age), ZAA
(Z at age), and
waa
(weight-at-age data).
NAA
looks ok.
therep$NAA
#> [,1] [,2] [,3] [,4] [,5] [,6]
#> [1,] 134809.00 11710.97 25945.08 17508.26 11476.00 13803.70
#> [2,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [3,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [4,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [5,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [6,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [7,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [8,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [9,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [10,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [11,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [12,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [13,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [14,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [15,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [16,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [17,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [18,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [19,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [20,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [21,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [22,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [23,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [24,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [25,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [26,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [27,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [28,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [29,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [30,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [31,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [32,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [33,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [34,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [35,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [36,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [37,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [38,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [39,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [40,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [41,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [42,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [43,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
#> [44,] 22026.47 22026.47 22026.47 22026.47 22026.47 22026.47
FAA
looks ok - no issues with F
or
selectivity.
therep$FAA[,1,] # middle dim is n.fleets = 1
#> [,1] [,2] [,3] [,4] [,5] [,6]
#> [1,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [2,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [3,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [4,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [5,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [6,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [7,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [8,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [9,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [10,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [11,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [12,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [13,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [14,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [15,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [16,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [17,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [18,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [19,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [20,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [21,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [22,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [23,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [24,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [25,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [26,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [27,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [28,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [29,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [30,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [31,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [32,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [33,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [34,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [35,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [36,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [37,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [38,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [39,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [40,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [41,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [42,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [43,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
#> [44,] 0.0501933 0.07552447 0.09057153 0.09692876 0.09922137 0.1
ZAA
looks ok - no issues with M
either.
therep$ZAA
#> [,1] [,2] [,3] [,4] [,5] [,6]
#> [1,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [2,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [3,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [4,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [5,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [6,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [7,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [8,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [9,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [10,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [11,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [12,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [13,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [14,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [15,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [16,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [17,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [18,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [19,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [20,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [21,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [22,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [23,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [24,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [25,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [26,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [27,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [28,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [29,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [30,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [31,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [32,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [33,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [34,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [35,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [36,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [37,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [38,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [39,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [40,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [41,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [42,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [43,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
#> [44,] 0.4551933 0.4115245 0.3865715 0.3719288 0.3552214 0.331099
Ah, here is the problem. The weight at age data has an entry of
-99
. This means that pred_catch
is negative on
line 880, and we take the log of a negative number on line 883.
input$data$waa[1,,]
#> [,1] [,2] [,3] [,4] [,5] [,6]
#> [1,] 0.210 0.296 0.348 0.374 0.382 0.428
#> [2,] 0.203 0.308 0.352 0.396 0.439 0.457
#> [3,] 0.218 0.289 0.376 0.432 0.435 0.481
#> [4,] 0.228 0.303 0.408 0.498 0.499 0.557
#> [5,] 0.215 0.283 0.381 0.504 0.513 0.542
#> [6,] 0.234 0.292 0.383 0.536 0.662 0.656
#> [7,] 0.189 0.301 0.364 0.475 0.590 0.662
#> [8,] 0.206 0.281 0.384 0.500 0.682 0.925
#> [9,] 0.140 0.262 0.342 0.474 0.596 0.650
#> [10,] 0.226 0.263 0.353 0.499 0.660 0.833
#> [11,] 0.175 0.261 0.339 0.496 0.668 0.819
#> [12,] 0.182 0.237 0.295 0.388 0.487 0.656
#> [13,] 0.183 0.260 0.365 0.408 0.504 0.608
#> [14,] 0.186 0.284 0.331 0.463 0.587 0.642
#> [15,] 0.247 0.268 0.353 0.404 0.520 0.631
#> [16,] 0.270 0.293 0.396 0.493 0.611 0.821
#> [17,] 0.061 0.216 0.275 0.489 0.735 0.957
#> [18,] 0.204 0.255 0.290 0.366 0.613 0.884
#> [19,] 0.090 0.214 0.294 0.378 0.664 0.798
#> [20,] 0.110 0.303 0.375 0.447 0.631 0.918
#> [21,] 0.122 0.314 0.420 0.439 0.640 1.040
#> [22,] 0.078 0.247 0.321 0.387 0.480 0.622
#> [23,] 0.076 0.216 0.325 0.401 0.579 0.758
#> [24,] 0.102 0.335 0.368 0.457 0.604 0.740
#> [25,] 0.139 0.251 0.396 0.466 0.584 0.768
#> [26,] 0.160 0.287 0.367 0.494 0.567 0.726
#> [27,] 0.131 0.309 0.400 0.557 0.629 1.695
#> [28,] 0.185 0.321 0.444 0.561 0.667 0.752
#> [29,] 0.145 0.360 0.419 0.567 0.684 0.824
#> [30,] 0.164 0.330 0.438 0.574 0.764 0.751
#> [31,] 0.095 0.313 0.413 0.572 0.722 0.945
#> [32,] 0.136 0.295 0.436 0.540 0.581 0.799
#> [33,] 0.102 0.295 0.415 0.511 0.634 0.795
#> [34,] 0.110 0.251 0.373 0.475 0.607 0.783
#> [35,] 0.111 0.268 0.363 0.472 0.628 0.834
#> [36,] 0.151 0.266 0.388 0.461 0.574 1.077
#> [37,] 0.105 0.281 0.367 0.502 0.601 0.753
#> [38,] 0.099 0.281 0.414 0.470 0.573 0.702
#> [39,] 0.130 0.280 0.412 0.491 0.572 0.717
#> [40,] 0.120 0.251 0.365 0.475 0.554 0.709
#> [41,] 0.154 0.254 0.351 0.453 0.565 0.683
#> [42,] -99.000 0.345 0.396 0.459 0.565 0.689
#> [43,] 0.046 0.281 0.388 0.476 0.548 0.685
#> [44,] 0.000 0.270 0.349 0.434 0.521 0.629
If we fix this issue with the data file, the model runs fine!