R Under development (unstable) (2024-07-02 r86866 ucrt) -- "Unsuffered Consequences" Copyright (C) 2024 The R Foundation for Statistical Computing Platform: x86_64-w64-mingw32/x64 R is free software and comes with ABSOLUTELY NO WARRANTY. You are welcome to redistribute it under certain conditions. Type 'license()' or 'licence()' for distribution details. R is a collaborative project with many contributors. Type 'contributors()' for more information and 'citation()' on how to cite R or R packages in publications. Type 'demo()' for some demos, 'help()' for on-line help, or 'help.start()' for an HTML browser interface to help. Type 'q()' to quit R. > library(magrittr) > library(unittest) > > library(gadget3) > > cmp_environment <- function (a, b) { + ordered_list <- function (x) { + x <- as.list(x) + # NB: Can't order an empty list + if (length(x) > 0) x[order(names(x))] else x + } + + ut_cmp_identical(ordered_list(a), ordered_list(b)) + } > > areas <- list(a=1, b=2, c=3, d=4) > stock_a <- g3_stock('stock_a', c(10)) %>% g3s_livesonareas(areas[c('a')]) > stock_ac <- g3_stock('stock_ac', c(10)) %>% g3s_livesonareas(unname(areas[c('a', 'c')])) # NB: Remove names so we generate defaults > stock_bcd <- g3_stock('stock_bcd', c(10)) %>% g3s_livesonareas(areas[c('b', 'c', 'd')]) > stock_aggregated <- g3_stock('stock_aggregated', c(10)) %>% g3s_areagroup(list(areas[c('b', 'c')], areas[c('d')])) > stock_1agg <- g3_stock('stock_1agg', c(10)) %>% g3s_areagroup(list(areas[c('b')], areas[c('c')])) > > ok(cmp_environment(stock_a$env, list( + stock__area = 1L, + stock__area_idx = stock_a$env$stock__area_idx, + stock__totalareas = 1L, + stock__areas = gadget3:::force_vector(a = 1L), + area_a = 1L, + stock__upperlen = Inf, + stock__minlen = gadget3:::force_vector("10:Inf" = 10), + stock__midlen = gadget3:::force_vector("10:Inf" = 10.5), + stock__maxmidlen = 10.5, + stock__minmidlen = 10.5, + stock__maxlen = gadget3:::force_vector("10:Inf" = Inf), + stock__plusdl = 1, + stock__dl = 1)), "stock_a: Environment populated with relevant areas") ok - stock_a: Environment populated with relevant areas > ok(cmp_environment(stock_ac$env, list( + stock__totalareas = 2L, + stock__areas = gadget3:::force_vector(area1 = 1L, area3 = 3L), + area_area1 = 1L, + area_area3 = 3L, + stock__upperlen = Inf, + stock__minlen = gadget3:::force_vector("10:Inf" = 10), + stock__midlen = gadget3:::force_vector("10:Inf" = 10.5), + stock__maxmidlen = 10.5, + stock__minmidlen = 10.5, + stock__maxlen = gadget3:::force_vector("10:Inf" = Inf), + stock__plusdl = 1, + stock__dl = 1)), "stock_c: Environment populated with default areas") ok - stock_c: Environment populated with default areas > ok(cmp_environment(stock_bcd$env, list( + stock__totalareas = 3L, + stock__areas = gadget3:::force_vector(b = 2L, c = 3L, d = 4L), + area_b = 2L, + area_c = 3L, + area_d = 4L, + stock__upperlen = Inf, + stock__minlen = gadget3:::force_vector("10:Inf" = 10), + stock__midlen = gadget3:::force_vector("10:Inf" = 10.5), + stock__maxmidlen = 10.5, + stock__minmidlen = 10.5, + stock__maxlen = gadget3:::force_vector("10:Inf" = Inf), + stock__plusdl = 1, + stock__dl = 1)), "stock_a: Environment populated with relevant areas") ok - stock_a: Environment populated with relevant areas > > nll <- 0.0 > stock_sum_a_ac <- 0.0 > stock_sum_ac_a <- 0.0 > stock_sum_ac_bcd <- 0.0 > stock_bcd__interacttotals <- g3_stock_instance(stock_bcd, 0) > stock_bcd_a_interactions <- 0L > stock_bcd_ac_interactions <- 0L > actions <- list( + # NB: livesonareas will add area names to environment, so this works + g3a_initialconditions(stock_a, ~(if (area == area_a) 1 else if (area == area_b) 2 else if (area == area_c) 3 else if (area == area_d) 4 else 0) * 100 + stock_a__minlen, ~0), + g3a_initialconditions(stock_ac, ~area * 1000 + stock_ac__minlen, ~0), + g3a_initialconditions(stock_bcd, ~area * 10000 + stock_bcd__minlen, ~0), + g3a_initialconditions(stock_aggregated, ~area * 1 + stock_bcd__minlen, ~0), + g3a_initialconditions(stock_1agg, ~area * 1 + stock_bcd__minlen, ~0), + list( + '5' = gadget3:::g3_step(~{ + comment("stock_sum_a_ac") + stock_iterate(stock_a, stock_intersect(stock_ac, { + stock_sum_a_ac <- stock_sum_a_ac + sum(stock_ss(stock_a__num)) + sum(stock_ss(stock_ac__num)) + })) + REPORT(stock_sum_a_ac) + + comment("stock_sum_ac_a") + stock_iterate(stock_ac, stock_intersect(stock_a, { + stock_sum_ac_a <- stock_sum_ac_a + sum(stock_ss(stock_ac__num)) + sum(stock_ss(stock_a__num)) + })) + REPORT(stock_sum_ac_a) + + comment("stock_sum_ac_bcd") + stock_iterate(stock_ac, stock_intersect(stock_bcd, { + stock_sum_ac_bcd <- stock_sum_ac_bcd + sum(stock_ss(stock_ac__num)) + sum(stock_ss(stock_bcd__num)) + })) + REPORT(stock_sum_ac_bcd) + + comment("stock_aggregated stock_a") + stock_iterate(stock_a, stock_intersect(stock_aggregated, { + stock_ss(stock_aggregated__num) <- + stock_ss(stock_aggregated__num) + + stock_ss(stock_a__num) + })) + + comment("stock_aggregated stock_ac") + stock_iterate(stock_ac, stock_intersect(stock_aggregated, { + stock_ss(stock_aggregated__num) <- + stock_ss(stock_aggregated__num) + + stock_ss(stock_ac__num) + })) + + comment("stock_aggregated stock_bcd") + stock_iterate(stock_bcd, stock_intersect(stock_aggregated, { + stock_ss(stock_aggregated__num) <- + stock_ss(stock_aggregated__num) + + stock_ss(stock_bcd__num) + })) + + comment("stock_1agg stock_a") + stock_iterate(stock_a, stock_intersect(stock_1agg, { + stock_ss(stock_1agg__num) <- + stock_ss(stock_1agg__num) + + stock_ss(stock_a__num) + })) + + comment("stock_1agg stock_ac") + stock_iterate(stock_ac, stock_intersect(stock_1agg, { + stock_ss(stock_1agg__num) <- + stock_ss(stock_1agg__num) + + stock_ss(stock_ac__num) + })) + + comment("stock_1agg stock_bcd") + stock_iterate(stock_bcd, stock_intersect(stock_1agg, { + stock_ss(stock_1agg__num) <- + stock_ss(stock_1agg__num) + + stock_ss(stock_bcd__num) + })) + + comment("interact bcd -> a") + stock_iterate(stock_bcd, stock_interact(stock_a, { + stock_ss(stock_bcd__interacttotals) <- stock_ss(stock_bcd__interacttotals) + stock_reshape(stock_bcd, stock_ss(stock_a__num)) + stock_bcd_a_interactions <- stock_bcd_a_interactions + area * 1000L + sub_area + }, prefix = "sub")) + comment("interact bcd -> ac") + stock_iterate(stock_bcd, stock_interact(stock_ac, { + stock_ss(stock_bcd__interacttotals) <- stock_ss(stock_bcd__interacttotals) + stock_reshape(stock_bcd, stock_ss(stock_ac__num)) + stock_bcd_ac_interactions <- stock_bcd_ac_interactions + area * 1000L + sub_area + }, prefix = "sub")) + }), + '999' = ~{ + REPORT(stock_a__num) + REPORT(stock_ac__num) + REPORT(stock_bcd__num) + REPORT(stock_aggregated__num) + REPORT(stock_1agg__num) + REPORT(stock_bcd_a_interactions) + REPORT(stock_bcd_ac_interactions) + REPORT(stock_bcd__interacttotals) + nll <- nll + g3_param('x', value = 1.0) + return(nll) + })) > model_fn <- g3_to_r(actions) > # model_fn <- edit(model_fn) > > params <- attr(model_fn, 'parameter_template') > result <- model_fn(params) > r <- attributes(result) > > # We iterated over the stock and populated using the area variable > ok(ut_cmp_identical( + r$stock_a__num, + array( + c(110), + dim = c(length = 1, area = 1), + dimnames = list(length = "10:Inf", area = "a"))), "stock_a__num populated, used names from areas lookup") ok - stock_a__num populated, used names from areas lookup > ok(ut_cmp_identical( + r$stock_ac__num, + array( + c(1010, 3010), + dim = c(length = 1, area = 2), + dimnames = list(length = "10:Inf", area = c("area1","area3")))), "stock_ac__num populated, generated default names") ok - stock_ac__num populated, generated default names > ok(ut_cmp_identical( + r$stock_bcd__num, + array( + c(20010, 30010, 40010), + dim = c(length = 1, area = 3), + dimnames = list(length = "10:Inf", area = c("b", "c", "d")))), "stock_bcd__num populated, used names from areas lookup") ok - stock_bcd__num populated, used names from areas lookup > > ok(ut_cmp_identical( + r$stock_aggregated__num, + array(c( + # NB: Areas b & c --> init + stock_ac + stock_bcd + (12) + (3010) + (20010 + 30010), + # NB: Area d --> init + stock_bcd + 14 + 40010), dim = c(length = 1, area = 2), dimnames = list(length = "10:Inf", area = c("area2", "area4")))), "stock_aggregated__num combination of all stocks") ok - stock_aggregated__num combination of all stocks > > ok(ut_cmp_identical( + r$stock_1agg__num, + array(c( + 12 + r$stock_bcd__num[1,'b'], + 13 + r$stock_ac__num[1,'area3'] + r$stock_bcd__num[1,'c']), dim = c(length = 1, area = 2), dimnames = list(length = "10:Inf", area = c("area2.b", "area3.c")))), "stock_1agg__num, got values for areas b & c") ok - stock_1agg__num, got values for areas b & c > > # Intersection works with any combination of single-area stock and multi-area stock > ok(ut_cmp_identical( + r$stock_sum_a_ac, + 110 + 1010), "stock_sum_a_ac: Only includes area a") ok - stock_sum_a_ac: Only includes area a > ok(ut_cmp_identical( + r$stock_sum_ac_a, + 110 + 1010), "stock_sum_ac_a: Only includes area a") ok - stock_sum_ac_a: Only includes area a > ok(ut_cmp_identical( + r$stock_sum_ac_bcd, + 3010 + 30010), "stock_sum_ac_bcd: Intersection is area c") ok - stock_sum_ac_bcd: Intersection is area c > > # Interaction is basically intersection for area > ok(ut_cmp_identical( + r$stock_bcd_a_interactions, + 0L), "stock_bcd_a_interactions: No interaction between bcd & a") ok - stock_bcd_a_interactions: No interaction between bcd & a > ok(ut_cmp_identical( + r$stock_bcd_ac_interactions, + 3003L), "stock_bcd_ac_interactions: Interact in area 3 (i.e. c") ok - stock_bcd_ac_interactions: Interact in area 3 (i.e. c > ok(ut_cmp_identical( + r$stock_bcd__interacttotals, + array( + c(0, 3010, 0), + dim = c(length = 1, area = 3), + dimnames = list(length = "10:Inf", area = c("b", "c", "d")))), "stock_bcd__interacttotals: Summed stock_ac__num in interaction") ok - stock_bcd__interacttotals: Summed stock_ac__num in interaction > > if (nzchar(Sys.getenv('G3_TEST_TMB'))) { + model_cpp <- g3_to_tmb(actions) + # model_cpp <- edit(model_cpp) + model_tmb <- g3_tmb_adfun(model_cpp, compile_flags = c("-O0", "-g")) + model_tmb_report <- model_tmb$report() + for (n in names(attributes(result))) { + ok(ut_cmp_equal( + as.vector(model_tmb_report[[n]]), + as.vector(attr(result, n)), + tolerance = 1e-5), paste("TMB and R match", n)) + } + } else { + writeLines("# skip: not running TMB tests") + } # skip: not running TMB tests > > proc.time() user system elapsed 0.70 0.07 0.78 # Looks like you passed all 14 tests.