library(tsDyn) suppressMessages(library(dplyr)) library(purrr) library(tidyr) select <- dplyr::select suppressWarnings(RNGversion("3.5.3")) ############################ ### Load data ############################ path_mod_multi <- system.file("inst/testdata/models_multivariate.rds", package = "tsDyn") if(path_mod_multi=="") path_mod_multi <- system.file("testdata/models_multivariate.rds", package = "tsDyn") models_multivariate <- readRDS(path_mod_multi) models_multivariate %>% mutate(across(where(is.list), class)) %>% as.data.frame() %>% head(12) ############################ ### VAR ############################ irf_any <- tsDyn:::irf_any irf_1 <- tsDyn:::irf_1 irf_1.nlVar <- tsDyn:::irf_1.nlVar ## manual comparisons mod_random_1 <- filter(models_multivariate, lag ==2)$object[[2]] mod_random_1_vars <- filter(models_multivariate, lag ==2)$object_vars[[2]] irf_any(mod_random_1, boot = FALSE)$irf[[1]] irf(mod_random_1, boot = FALSE)$irf[[1]] irf(mod_random_1_vars, boot = FALSE)$irf[[1]] irf_any(mod_random_1, boot = FALSE, ortho = FALSE)$irf[[1]] irf(mod_random_1, boot = FALSE, ortho = FALSE)$irf[[1]] irf(mod_random_1_vars, boot = FALSE, ortho = FALSE)$irf[[1]] ### irf _1 models_IRF_1 <- models_multivariate %>% filter(model == "VAR") %>% mutate(irf = map(object, ~irf_1(.))) models_IRF_1$irf %>% bind_rows() %>% head() %>% print(digits=3) ### irf_any # irf.NULL <- function(x) NULL # irf.ca.jo <- function(x) irf(vec2var(ca.jo)) models_VAR <- models_multivariate %>% filter(model == "VAR") ## older method models_IRF_any <- models_multivariate %>% filter(model == "VAR") %>% mutate(ortho = list(tibble(ortho =c(TRUE, FALSE)))) %>% unnest(., ortho) %>% mutate(irf = map2(object, ortho, ~irf_any(.x, boot = TRUE, runs = 1, seed = 7, ortho = .y)), irf_vars = map2(object_vars, ortho, ~irf(.x, runs = 1, seed = 7, ortho = .y)), irf_vec2 = map2(object, ortho, ~irf(.x, boot = FALSE, runs = 1, seed = 7, ortho = .y))) models_IRF_any %>% mutate(across(where(is.list), class)) %>% as.data.frame() ## showquick summary irf_extract_here <- function(x) { head(x$irf[[1]], 2) %>% as.data.frame() %>% mutate(type = "irf") %>% rbind(head(x$Upper[[1]], 2) %>% as.data.frame() %>% mutate(type = "Upper_CI")) %>% relocate(type) } ## show head of irf any map_dfr(models_IRF_any$irf, irf_extract_here) %>% as.data.frame() %>% head(10)%>% mutate(across(where(is.numeric), ~round(., 6))) ## compare with vars all.equal(models_IRF_any$irf[[1]]$irf, models_IRF_any$irf_vars[[1]]$irf) models_IRF_any$irf[[1]]$irf[[1]] models_IRF_any$irf_vars[[1]]$irf[[1]] models_IRF_any$irf_vec2[[1]]$irf[[1]] comp <- models_IRF_any %>% mutate(comp_irf_tsD_vars = map2(irf, irf_vars, ~all.equal(.x$irf, .y$irf)), is_same = map_lgl(comp_irf_tsD_vars, ~isTRUE(.)), comp_irf_tsDOld_vars = map2(irf_vec2, irf_vars, ~all.equal(.x$irf, .y$irf)), is_same_tssDvec2 = map_lgl(comp_irf_tsDOld_vars, ~isTRUE(.)), comp_irf_tsDOld_tsDNew = map2_lgl(irf, irf_vec2, ~all.equal(.x$irf, .y$irf)), is_same_tsD_2ver = map_lgl(comp_irf_tsDOld_tsDNew, ~isTRUE(.))) %>% dplyr::select(-starts_with("irf"), -starts_with("comp_irf"), comp_irf_tsDOld_tsDNew) comp %>% dplyr::select(-starts_with("object")) %>% as.data.frame() ############################ ### VECM ############################ models_VECM <- models_multivariate %>% filter(model == "VECM") %>% mutate(irf = map(object, ~irf_any(., boot = TRUE, runs = 1, seed = 7, ortho = FALSE))) ## show two first of first componment models_VECM %>% mutate(irf = map(irf, irf_extract_here)) %>% dplyr::select(-object, -object_vars) %>% unnest(irf) %>% as.data.frame() %>% mutate(across(where(is.numeric), ~round(., 6))) ## plot 1 plot(models_VECM$irf[[1]]) ############################ ### TVAR ############################ models_TVAR <- models_multivariate %>% filter(model == "TVAR") ## test 1 tvar_1 <- models_TVAR$object[[1]] irf(tvar_1, runs = 2, seed = 123) ## regime specific for TVAR models_TVAR_irf <- models_TVAR %>% mutate(irf_L = map(object, ~irf_any(., boot = TRUE, runs = 1, seed = 7, ortho = FALSE, regime = "L"))) ## show two first of first componment models_TVAR_irf %>% mutate(irf = map(irf_L, irf_extract_here)) %>% dplyr::select(-object, -object_vars, -irf_L ) %>% unnest(irf) %>% as.data.frame() %>% mutate(across(where(is.numeric), ~round(., 6))) ## plot 1 plot(models_TVAR_irf$irf_L[[1]]) ############################ ### TVECM ############################ models_TVECM <- models_multivariate %>% filter(model == "TVECM") ## test 1 tvecm_1 <- models_TVECM$object[[1]] tsDyn:::irf_1(x=tvecm_1 , n.ahead = 10, cumulative = FALSE, regime = "L", ortho = TRUE) tsDyn:::irf_1(x=tvecm_1 , n.ahead = 10, cumulative = FALSE, regime = "L", ortho = FALSE) irf(x=tvecm_1, runs = 2, seed = 123) ## regime specific for TVECM models_TVECM_irf <- models_TVECM %>% mutate(irf_L = map(object, ~suppressWarnings(irf_any(., boot = TRUE, runs = 1, seed = 7, ortho = FALSE, regime = "L")))) ## show two first of first componment models_TVECM_irf %>% mutate(irf = map(irf_L, irf_extract_here)) %>% select(-object, -object_vars, -irf_L ) %>% unnest(irf) %>% as.data.frame() %>% mutate(across(where(is.numeric), ~round(., 6))) ## plot 1 plot(models_TVECM_irf$irf_L[[1]])