Commit 72ae463d authored by matbuoro's avatar matbuoro
Browse files

Mise à jour estimations 2016

parent 0f8ba986
......@@ -12,12 +12,13 @@ library(mcmcplots)
##-----------------------------INFO ----------------------------------##
year <- "2015"
year <- "2016"
site <- "Nivelle"
stade <- "adult"
## WORKING DIRECTORY:
# work.dir<-paste("~/Documents/RESEARCH/PROJECTS/ORE/Abundance",site,stade,sep="/")
# work.dir<-paste("/media/ORE/Abundance",site,stade,sep="/")
# setwd(work.dir)
......@@ -53,8 +54,8 @@ filename <- file.path(work.dir, model)
#---------------------------ANALYSIS-----------------------------##
nChains = length(inits) # Number of chains to run.
adaptSteps = 1000 # Number of steps to "tune" the samplers.
nburnin=500 # Number of steps to "burn-in" the samplers.
nstore=1000 # Total number of steps in chains to save.
nburnin=2000 # Number of steps to "burn-in" the samplers.
nstore=5000 # Total number of steps in chains to save.
nthin=1 # Number of steps to "thin" (1=keep every step).
#nPerChain = ceiling( ( numSavedSteps * thinSteps ) / nChains ) # Steps per chain.
......
This diff is collapsed.
......@@ -4,7 +4,7 @@ modelCompile(1)
modelSetRN(1)
modelInits('/home/basp-meco88/Documents/RESEARCH/PROJECTS/ORE/Abundance/Nivelle/adult/bugs/inits1.txt',1)
modelGenInits()
modelUpdate(500,1,500)
modelUpdate(2000,1,2000)
samplesSet(mup_11_1)
samplesSet(sigmap_11_1)
samplesSet(mup_11_2)
......@@ -139,7 +139,7 @@ summarySet(c_3SW)
summarySet(c_tot)
summarySet(P_1SW)
summarySet(P_MSW)
modelUpdate(1000,1,1000)
modelUpdate(5000,1,5000)
samplesCoda('*', '/home/basp-meco88/Documents/RESEARCH/PROJECTS/ORE/Abundance/Nivelle/adult/bugs//')
summaryStats('*')
modelQuit('y')
=============================
DIAGNOSTICS
=============================
---------------------------
Heidelberger and Welch's convergence diagnostic
heidel.diag is a run length control diagnostic based on a criterion of relative accuracy for the estimate of the mean. The default setting corresponds to a relative accuracy of two significant digits.
heidel.diag also implements a convergence diagnostic, and removes up to half the chain in order to ensure that the means are estimated from a chain that has converged.
Stationarity start p-value
test iteration
mup_11_1 passed 1 0.2294
sigmap_11_1 passed 1 0.1678
mup_11_2 passed 1 0.4416
sigmap_11_2 failed NA 0.0198
mupi_EF passed 1 0.1983
sigmapi_EF passed 1 0.1726
mup_21 passed 1 0.6430
sigmap_21 passed 1 0.1533
k_1 passed 1 0.2935
k_2 passed 1 0.7129
eta_1 passed 1 0.1334
eta_2 passed 1 0.2324
rho passed 1 0.6888
shape_lambda passed 1 0.2802
rate_lambda passed 1 0.3441
lambda_tot0 passed 1 0.8005
a_1.1SW passed 1 0.8186
a_2.1SW passed 1 0.6540
Halfwidth Mean Halfwidth
test
mup_11_1 passed 0.2447 0.005560
sigmap_11_1 failed 0.4376 0.068686
mup_11_2 passed 0.1204 0.002275
sigmap_11_2 <NA> NA NA
mupi_EF passed 0.2476 0.002445
sigmapi_EF passed 0.7760 0.023879
mup_21 passed 0.6521 0.002417
sigmap_21 passed 0.5789 0.015989
k_1 passed 1.1704 0.064811
k_2 passed 2.4299 0.160761
eta_1 passed 3.3534 0.062855
eta_2 passed 5.1254 0.117029
rho passed 0.9452 0.009248
shape_lambda passed 2.8566 0.068897
rate_lambda passed 0.0151 0.000405
lambda_tot0 passed 142.0949 4.915053
a_1.1SW passed 10.0370 0.246490
a_2.1SW passed 2.9461 0.073931
---------------------------
Geweke's convergence diagnostic
Geweke (1992) proposed a convergence diagnostic for Markov chains based on a test for equality of the means of the first and last part of a Markov chain (by default the first 10% and the last 50%).
If the samples are drawn from the stationary distribution of the chain, the two means are equal and Geweke's statistic has an asymptotically standard normal distribution.
The test statistic is a standard Z-score: the difference between the two sample means divided by its estimated standard error. The standard error is estimated from the spectral density at zero and so takes into account any autocorrelation.
The Z-score is calculated under the assumption that the two parts of the chain are asymptotically independent, which requires that the sum of frac1 and frac2 be strictly less than 1.
Fraction in 1st window = 0.1
Fraction in 2nd window = 0.5
mup_11_1 sigmap_11_1 mup_11_2 sigmap_11_2 mupi_EF sigmapi_EF mup_21 sigmap_21 k_1
-0.6974 -2.6615 -0.3153 0.6576 -2.2736 -0.2349 0.3858 1.2018 0.5753
k_2 eta_1 eta_2 rho shape_lambda rate_lambda lambda_tot0 a_1.1SW a_2.1SW
-0.0455 1.1900 1.4736 -0.5396 -0.5237 -0.5010 -0.5615 0.8885 0.8970
---------------------------
Raftery and Lewis's diagnostic
Quantile (q) = 0.025
Accuracy (r) = +/- 0.005
Probability (s) = 0.95
Burn-in Total Lower bound Dependence
(M) (N) (Nmin) factor (I)
mup_11_1 4 5124 3746 1.37
sigmap_11_1 56 55960 3746 14.90
mup_11_2 10 11628 3746 3.10
sigmap_11_2 60 50640 3746 13.50
mupi_EF 3 4338 3746 1.16
sigmapi_EF 12 13334 3746 3.56
mup_21 4 5062 3746 1.35
sigmap_21 22 19268 3746 5.14
k_1 36 40242 3746 10.70
k_2 58 61786 3746 16.50
eta_1 4 4816 3746 1.29
eta_2 4 5038 3746 1.34
rho 44 48640 3746 13.00
shape_lambda 9 9892 3746 2.64
rate_lambda 12 13658 3746 3.65
lambda_tot0 5 5973 3746 1.59
a_1.1SW 10 12194 3746 3.26
a_2.1SW 9 9483 3746 2.53
This diff is collapsed.
Supports Markdown
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment