Add multivariate crps learning slides

25_07_phd_defense/assets/01_common.R

@@ -1,4 +1,5 @@
 text_size <- 16
+linesize <- 1
 width <- 12
 height <- 6
 
@@ -29,3 +30,26 @@ lamgrid <- c(-Inf, 2^(-15:25))
 
 # Gamma grid
 gammagrid <- sort(1 - sqrt(seq(0, 0.99, .05)))
+
+material_pals <- c(
+  "red", "pink", "purple", "deep-purple", "indigo",
+  "blue", "light-blue", "cyan", "teal", "green", "light-green", "lime",
+  "yellow", "amber", "orange", "deep-orange", "brown", "grey", "blue-grey"
+)
+cols <- purrr::map(material_pals, ~ ggsci::pal_material(.x)(10)) %>%
+  purrr::reduce(cbind)
+colnames(cols) <- material_pals
+
+cols %>%
+  as_tibble() %>%
+  mutate(idx = as.factor(1:10)) %>%
+  pivot_longer(-idx, names_to = "var", values_to = "val") %>%
+  mutate(var = factor(var, levels = material_pals[19:1])) %>%
+  ggplot() +
+  xlab(NULL) +
+  ylab(NULL) +
+  geom_tile(aes(x = idx, y = var, fill = val)) +
+  scale_fill_identity() +
+  scale_x_discrete(expand = c(0, 0)) +
+  scale_y_discrete(expand = c(0, 0)) +
+  theme_minimal() -> plot_cols

25_07_phd_defense/assets/library.bib

@@ -14,6 +14,16 @@
   booktitle = {Oxford Research Encyclopedia of Economics and Finance},
   year      = {2019}
 }
+@article{marcjasz2022distributional,
+  title     = {Distributional neural networks for electricity price forecasting},
+  author    = {Marcjasz, Grzegorz and Narajewski, Micha{\l} and Weron, Rafa{\l} and Ziel, Florian},
+  journal   = {Energy Economics},
+  volume    = {125},
+  pages     = {106843},
+  year      = {2023},
+  doi       = {10.1016/j.eneco.2023.106843},
+  publisher = {Elsevier}
+}
 @article{atiya2020does,
   title     = {Why does forecast combination work so well?},
   author    = {Atiya, Amir F},