Fixed internals for calling negative binomial with string "nbinom" (i…

…ssue #2), fixed documentation of reconc_BUIS and of reconc_MCMC with correct names of parameters for negative binomial (issue #2). Added a warning when size=0 in negative binomial (issue #4). Added a test for negative binomial reconciliation from parametric distribution.
IDSIA · Nov 17, 2023 · a0570bb · a0570bb
1 parent 4715f43
commit a0570bb
Show file tree

Hide file tree

Showing 7 changed files with 79 additions and 6 deletions.
diff --git a/R/reconc.R b/R/reconc.R
@@ -8,8 +8,9 @@
       samples = stats::rnorm(n=n, mean = params[[1]], sd = params[[2]]) },
     "poisson"  = {
       samples = stats::rpois(n=n, lambda = params[[1]]) },
-    "negbin"   = {
+    "nbinom"   = {
       samples <-if (params[[2]] == 0) {
+          warning("Parameter size=0 in nbinom, using a Poisson with lambda=mu.")
           stats::rpois(n=n, lambda = params[[1]])
         } else {
           stats::rnbinom(n=n, mu = params[[1]], size = params[[2]])
@@ -24,7 +25,7 @@
       pmf = stats::dnorm(x=x, mean = params[[1]], sd = params[[2]]) },
     "poisson"  = {
       pmf = stats::dpois(x=x, lambda = params[[1]]) },
-    "negbin"   = {
+    "nbinom"   = {
       pmf = stats::dnbinom(x=x, mu = params[[1]], size = params[[2]]) },
   )
   return(pmf)
@@ -87,7 +88,7 @@
 #'
 #' * mean and sd for the Gaussian base forecast, see \link[stats]{Normal}, if `distr`='gaussian';
 #' * lambda for the Poisson base forecast, see \link[stats]{Poisson}, if `distr`='poisson';
-#' * size and probability of success for the negative binomial base forecast, see \link[stats]{NegBinomial}, if `distr`='nbinom'.
+#' * mu and size for the negative binomial base forecast, see \link[stats]{NegBinomial}, if `distr`='nbinom'.
 #'
 #' The order of the `base_forecast` list is given by the order of the time series in the summing matrix.
 #'

diff --git a/R/reconc_MH.R b/R/reconc_MH.R
@@ -29,7 +29,7 @@
 #'
 #' * mean and sd for the Gaussian base forecast, see \link[stats]{Normal}, if `distr`='gaussian';
 #' * lambda for the Poisson base forecast, see \link[stats]{Poisson}, if `distr`='poisson';
-#' * size and probability of success for the negative binomial base forecast, see \link[stats]{NegBinomial}, if `distr`='nbinom'.
+#' * mu and size for the negative binomial base forecast, see \link[stats]{NegBinomial}, if `distr`='nbinom'.
 #'
 #' The order of the `base_forecast` list is given by the order of the time series in the summing matrix.
 #'

diff --git a/man/reconc_BUIS.Rd b/man/reconc_BUIS.Rd
diff --git a/man/reconc_MCMC.Rd b/man/reconc_MCMC.Rd
diff --git a/tests/testthat/dataForTests/Monthly-NegBin_S.csv b/tests/testthat/dataForTests/Monthly-NegBin_S.csv
@@ -0,0 +1,28 @@
+1.000000000000000000e+00,1.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00
+0.000000000000000000e+00,0.000000000000000000e+00,1.000000000000000000e+00,1.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00
+0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,1.000000000000000000e+00,1.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00
+0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,1.000000000000000000e+00,1.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00
+0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,1.000000000000000000e+00,1.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00
+0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,1.000000000000000000e+00,1.000000000000000000e+00
+1.000000000000000000e+00,1.000000000000000000e+00,1.000000000000000000e+00,1.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00
+0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,1.000000000000000000e+00,1.000000000000000000e+00,1.000000000000000000e+00,1.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00
+0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,1.000000000000000000e+00,1.000000000000000000e+00,1.000000000000000000e+00,1.000000000000000000e+00
+1.000000000000000000e+00,1.000000000000000000e+00,1.000000000000000000e+00,1.000000000000000000e+00,1.000000000000000000e+00,1.000000000000000000e+00,1.000000000000000000e+00,1.000000000000000000e+00,1.000000000000000000e+00,1.000000000000000000e+00,1.000000000000000000e+00,1.000000000000000000e+00
+1.000000000000000000e+00,1.000000000000000000e+00,1.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00
+0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,1.000000000000000000e+00,1.000000000000000000e+00,1.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00
+0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,1.000000000000000000e+00,1.000000000000000000e+00,1.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00
+0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,1.000000000000000000e+00,1.000000000000000000e+00,1.000000000000000000e+00
+1.000000000000000000e+00,1.000000000000000000e+00,1.000000000000000000e+00,1.000000000000000000e+00,1.000000000000000000e+00,1.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00
+0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,1.000000000000000000e+00,1.000000000000000000e+00,1.000000000000000000e+00,1.000000000000000000e+00,1.000000000000000000e+00,1.000000000000000000e+00
+1.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00
+0.000000000000000000e+00,1.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00
+0.000000000000000000e+00,0.000000000000000000e+00,1.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00
+0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,1.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00
+0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,1.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00
+0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,1.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00
+0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,1.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00
+0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,1.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00
+0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,1.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00
+0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,1.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00
+0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,1.000000000000000000e+00,0.000000000000000000e+00
+0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,0.000000000000000000e+00,1.000000000000000000e+00
diff --git a/tests/testthat/dataForTests/Monthly-NegBin_basef.csv b/tests/testthat/dataForTests/Monthly-NegBin_basef.csv
@@ -0,0 +1,28 @@
+21,21
+17,17
+17,17
+16,16
+18,18
+17,17
+38,38
+33,33
+36,36
+107,107
+28,28
+27,27
+25,25
+27,27
+56,56
+52,52
+9.57403021748178,9.57403021748178
+9.68537706648931,9.68537706648931
+6.43069767393172,6.43069767393172
+9.15223813033663,9.15223813033663
+8.20872759446502,8.20872759446502
+7.59547974565066,7.59547974565066
+8.68294157320634,8.68294157320634
+5.67333298618905,5.67333298618905
+8.28496145200916,8.28496145200916
+8.52532392018475,8.52532392018475
+7.2887088812422,7.2887088812422
+8.59556125826202,8.59556125826202
diff --git a/tests/testthat/test-examples.R b/tests/testthat/test-examples.R
@@ -47,6 +47,22 @@ test_that("Monthly, in_type=='params', distr='poisson'",{
   expect_no_error(res.buis)
 })
 
+
+test_that("Monthly, in_type=='params', distr='nbinom'",{
+  S = read.csv(file = "dataForTests/Monthly-NegBin_S.csv", header = FALSE)
+  S = as.matrix(S)
+  base_forecasts_in = read.csv(file = "dataForTests/Monthly-NegBin_basef.csv", header = FALSE)
+  base_forecasts = list()
+  for (i in 1:nrow(base_forecasts_in)) {
+    base_forecasts[[i]] = list(as.numeric(base_forecasts_in[i,]))[[1]]
+  }
+  res.buis = reconc_BUIS(S, base_forecasts,
+                         in_type = "params", distr = "nbinom", num_samples = 10000, seed=42)
+  expect_no_error(res.buis)
+})
+
+
+
 test_that("Monthly, in_type=='samples', distr='continuous'",{
   # Run IS Reconc from samples
   S = read.csv(file = "dataForTests/Monthly-Gaussian_S.csv", header = FALSE)