This document follows the modelling pipeline to produce a list of introduced species ranked by emerging status.
Load libraries:
library(tidyverse) # To do data science
library(tidylog) # To provide feedback on dplyr functions
library(here) # To find filesWe read the emerging status based on GAM, saved as one of the outputs of modelling pipeline. We do it for Belgium, its regions (Flanders, Wallonia and Brussels) and protected areas and for both occurrences and observed occupancy (number of occupied cells):
regions <- list(
"Belgium",
"Flanders",
"Wallonia",
"Brussels",
"natura2000"
)
names(regions) <- regions
indicators <- c("occs", "occupancy")
names(indicators) <- indicators
# Create all combinations regions/indicators
combinations <- expand_grid(region = regions, indicator = indicators)
# Apply a function to read files for each combination
em_gam_list <- pmap(combinations, function(region, indicator) {
paste("Reading", indicator, "for", region)
readr::read_tsv(
file = here::here(
"data",
"output",
"GAM_outputs",
paste0("summary_GAM_", indicator, "_", region, ".tsv")
),
na = ""
) %>%
tidylog::mutate(
indicator = paste(indicator, region, sep = "_"),
model = "GAM"
)
})
# Name the results
names(em_gam_list) <- paste("gam", combinations$indicator,
combinations$region,
sep = "_"
)Preview GAM ouput for modelling the number of occurrences in Belgium (100 rows):
em_gam_list$gam_occs_Belgium %>% head(100)Preview GAM ouput for modelling occupancy (number of 1x1km cells) in Belgium (100 rows):
em_gam_list$gam_occupancy_Belgium %>% head(100)The emerging status (column em_status) score is encoded
as follows:
As you can see in preview above, GAM could not be applied to all taxa
(em_status = NA). Number of taxon/year
combinations without emerging status:
em_gam_na <- purrr::map(
em_gam_list,
function(df) {
nas <- df %>%
dplyr::filter(is.na(em_status)) %>%
nrow()
total <- nrow(df)
list("NAs" = nas, "Total" = total)
}
)
purrr::iwalk(
em_gam_na,
function(x, y) {
message(paste0("Number of NAs (", y, "): ", x$NAs, "/", x$Total))
}
)In these cases we will use the emerging status as assessed by applying decision rules.
We read the emerging status based on decision rules, saved as one of the outputs of modelling pipeline:
# Apply a function to read files for each combination
em_decision_rules_list <- pmap(combinations, function(region, indicator) {
paste("Reading", indicator, "for", region)
readr::read_tsv(
file = here::here(
"data",
"output",
"decision_rules_outputs",
paste0("output_decision_rules_", indicator, "_", region, ".tsv")
),
na = ""
) %>%
tidylog::mutate(
indicator = paste(indicator, region, sep = "_"),
model = "decision_rules"
)
})
# Name the results
names(em_decision_rules_list) <- paste(
"decision_rules",
combinations$indicator,
combinations$region,
sep = "_"
)Previewing output of decision rules for occupancy in Natura2000 protected areas (100 rows):
em_decision_rules_list$decision_rules_occupancy_natura2000 %>% head(100)Read list of scientific names:
taxa_names <- readr::read_tsv(
file = here::here(
"data",
"interim",
"timeseries_taxonomic_info.tsv"
),
na = ""
)These names will be added to ranking for better readability.
Merge results from GAM and decision rules:
em_list <- append(em_gam_list, em_decision_rules_list)We have now to make a distinction between Belgium and its provinces. Assessment at Belgian level takes into account the emerging status of the protected areas as well, while at regional level this is not (yet) done.
However, We can define a general function to spread columns for ranking. We spread the data frames by model (GAM or decision rules), year and indicator (occurrences or occupancy). We then assign a unique emerging score (if GAM is NA, we choose decision rules). We spread by years and indicators and add scientific names.
spread_em_dfs <- function(em_list) {
# Make a data.frame binding all dfs by row
em_df <-
dplyr::bind_rows(em_list) %>%
tidylog::select(-c(
dplyr::starts_with("growth"),
dplyr::starts_with("dr")
)) %>%
# Spread by model
tidylog::pivot_wider(
names_from = model,
id_cols = c(taxonKey, year, indicator),
values_from = em_status
) %>%
# Assign unique emerging score (if GAM is NA, choose decision rules)
tidylog::mutate(
em = ifelse(is.na(GAM), decision_rules, GAM),
method = ifelse(is.na(GAM), "decision_rules", "GAM")
) %>%
# Spread by years
tidylog::pivot_wider(
names_from = year,
values_from = em,
id_cols = c(taxonKey, indicator),
names_prefix = paste0("year_")
) %>%
# Spread by indicator
tidylog::pivot_wider(
names_from = indicator,
id_cols = taxonKey,
values_from = starts_with("year"),
names_prefix = "em_status_"
) %>%
# Add scientific names
tidylog::left_join(taxa_names, by = "taxonKey")
em_df
}Select only list elements containing Belgium or
natura2000 in the name:
em_list_be <- em_list[grep("Belgium|natura2000", names(em_list))]Apply the function spread_em_dfs() to Belgium:
em_df_be <- spread_em_dfs(em_list_be)To rank taxa equally placed, we use the average of the minimal guaranteed growth of number of occurrences in all Belgium over all the evaluation years.
# Add mean growth to em_df
em_df_be <-
em_df_be %>%
tidylog::left_join(em_gam_list$gam_occs_Belgium %>%
tidylog::group_by(taxonKey) %>%
tidylog::summarize(mean_growth = mean(growth, na.rm = TRUE))) %>%
tidylog::relocate(.data$mean_growth, .before = .data$canonicalName)A preview (100 rows):
em_df_be %>%
tidylog::filter(!is.na(mean_growth)) %>%
tidylog::slice_head(n = 100)Select only list elements containing the region in the name:
regions <- list("Flanders", "Wallonia", "Brussels")
names(regions) <- regions
em_list_regions <- purrr::map(
regions,
function(r) em_list[grep(r, names(em_list))]
)Apply the function spread_em_dfs() to each region:
em_df_regions <- purrr::map(em_list_regions, spread_em_dfs)To rank taxa equally placed, we use the average of the minimal guaranteed growth of number of occurrences over all the evaluation years.
# Add mean growth to em_df
em_df_regions <-
em_df_regions %>% purrr::imap(
function(df, region) {
occs_column <- paste0("gam_occs_", region)
df %>%
tidylog::left_join(em_gam_list[[occs_column]] %>%
tidylog::group_by(taxonKey) %>%
tidylog::summarize(mean_growth = mean(growth, na.rm = TRUE))) %>%
tidylog::relocate(.data$mean_growth, .before = .data$canonicalName)
}
)A preview (100 rows) for each region:
em_df_regions %>%
purrr::map(function(df) {
df %>%
tidylog::filter(!is.na(mean_growth)) %>%
tidylog::slice_head(n = 100)
})## $Flanders
## # A tibble: 100 × 13
## taxonKey year_2020_em_status_…¹ year_2020_em_status_…² year_2021_em_status_…³
## <dbl> <dbl> <dbl> <dbl>
## 1 1031394 0 1 0
## 2 1031742 1 1 1
## 3 1047536 3 3 3
## 4 1309600 2 2 3
## 5 1421091 3 0 3
## 6 1652150 1 1 1
## 7 1690429 0 1 1
## 8 1718308 0 1 0
## 9 1749449 3 2 3
## 10 1749580 0 2 0
## # ℹ 90 more rows
## # ℹ abbreviated names: ¹year_2020_em_status_occs_Flanders,
## # ²year_2020_em_status_occupancy_Flanders, ³year_2021_em_status_occs_Flanders
## # ℹ 9 more variables: year_2021_em_status_occupancy_Flanders <dbl>,
## # year_2022_em_status_occs_Flanders <dbl>,
## # year_2022_em_status_occupancy_Flanders <dbl>, mean_growth <dbl>,
## # canonicalName <chr>, kingdomKey <dbl>, classKey <dbl>, class <chr>, …
##
## $Wallonia
## # A tibble: 100 × 13
## taxonKey year_2020_em_status_…¹ year_2020_em_status_…² year_2021_em_status_…³
## <dbl> <dbl> <dbl> <dbl>
## 1 1047536 3 0 3
## 2 1690429 0 0 0
## 3 1718308 1 3 1
## 4 2020655 0 2 0
## 5 2041682 1 3 1
## 6 2226990 3 3 3
## 7 2337607 1 0 1
## 8 2362868 1 1 1
## 9 2379089 0 0 0
## 10 2379245 0 1 0
## # ℹ 90 more rows
## # ℹ abbreviated names: ¹year_2020_em_status_occs_Wallonia,
## # ²year_2020_em_status_occupancy_Wallonia, ³year_2021_em_status_occs_Wallonia
## # ℹ 9 more variables: year_2021_em_status_occupancy_Wallonia <dbl>,
## # year_2022_em_status_occs_Wallonia <dbl>,
## # year_2022_em_status_occupancy_Wallonia <dbl>, mean_growth <dbl>,
## # canonicalName <chr>, kingdomKey <dbl>, classKey <dbl>, class <chr>, …
##
## $Brussels
## # A tibble: 80 × 13
## taxonKey year_2020_em_status_…¹ year_2020_em_status_…² year_2021_em_status_…³
## <dbl> <dbl> <dbl> <dbl>
## 1 1047536 3 3 3
## 2 1690429 1 2 1
## 3 2020655 1 2 1
## 4 2037925 3 3 3
## 5 2041682 3 3 3
## 6 2078852 3 1 1
## 7 2436940 1 2 1
## 8 2437450 0 2 0
## 9 2439261 2 2 2
## 10 2479226 3 1 3
## # ℹ 70 more rows
## # ℹ abbreviated names: ¹year_2020_em_status_occs_Brussels,
## # ²year_2020_em_status_occupancy_Brussels, ³year_2021_em_status_occs_Brussels
## # ℹ 9 more variables: year_2021_em_status_occupancy_Brussels <dbl>,
## # year_2022_em_status_occs_Brussels <dbl>,
## # year_2022_em_status_occupancy_Brussels <dbl>, mean_growth <dbl>,
## # canonicalName <chr>, kingdomKey <dbl>, classKey <dbl>, class <chr>, …Now we are ready to rank the alien species based on the emerging status.
The ranking is based on the highest emerging status. The following priority rules are applied, in order of importance:
We first define a function to rank the emerging status of species
based on the priority rules above. The function takes as input the
years, indicators, regions and the data frame with emerging status
scores. It returns a data frame with the ranking of species based on the
priority rules. The function is called
hierarchical_ranking() and is defined as follows:
hierarchical_ranking <- function(years, indicators, regions, df) {
# Generate the column names dynamically
cols_to_sort <- expand.grid(indicator = indicators, region = regions, year = years) %>%
mutate(col_name = paste("year", year, "em_status", indicator, region, sep = "_")) %>%
pull(col_name)
# Add mean_growth to the end of the sorting columns
cols_to_sort <- c(cols_to_sort, "mean_growth")
# Apply the sorting
ranking <- df %>%
arrange(across(all_of(cols_to_sort), desc)) %>%
tidylog::select(
taxonKey, canonicalName, kingdom, class,
cols_to_sort,
dplyr::everything()
)
ranking
}We apply the function to Belgium and its Natura2000 protected areas:
years <- sort(
unique(
as.numeric(gsub("^year_(\\d+)_em_status_.*", "\\1", names(em_df_be)))
),
decreasing = TRUE
)
indicators <- c("occupancy", "occs")
regions <- c("natura2000", "Belgium")
ranking_be <- hierarchical_ranking(years, indicators, regions, em_df_be)
ranking_beWe have not done (yet) an analysis for protected areas at regional level. So, the ranking for regions will be simplified. It means that the priority rule 2 described above will not be applied. The priority rules will then look as follows:
We apply the function hierarchical_ranking() to
Flanders, Wallonia and Brussels:
years <- sort(
unique(
as.numeric(gsub("^year_(\\d+)_em_status_.*", "\\1", names(em_df_be)))
),
decreasing = TRUE
)
indicators <- c("occupancy", "occs")
regions <- c("Flanders", "Wallonia", "Brussels")
ranking_regions <- purrr::imap(
em_df_regions,
function(em_df, region) hierarchical_ranking(years, indicators, region, em_df)
)
ranking_regions## $Flanders
## # A tibble: 1,655 × 13
## taxonKey canonicalName kingdom class year_2022_em_status_…¹
## <dbl> <chr> <chr> <chr> <dbl>
## 1 6098843 Ctenolepisma longicaudatum Animalia Insecta 3
## 2 2225772 Hemigrapsus sanguineus Animalia Malacost… 3
## 3 1047536 Leptinotarsa decemlineata Animalia Insecta 3
## 4 2498344 Cygnus atratus Animalia Aves 3
## 5 3098912 Cosmos bipinnatus Plantae Magnolio… 3
## 6 5232437 Branta canadensis Animalia Aves 3
## 7 2479226 Psittacula krameri Animalia Aves 3
## 8 2495494 Geopelia cuneata Animalia Aves 3
## 9 2650827 Cyrtomium fortunei Plantae Polypodi… 3
## 10 6096602 Lophocolea semiteres Plantae Jungerma… 3
## # ℹ 1,645 more rows
## # ℹ abbreviated name: ¹year_2022_em_status_occupancy_Flanders
## # ℹ 8 more variables: year_2022_em_status_occs_Flanders <dbl>,
## # year_2021_em_status_occupancy_Flanders <dbl>,
## # year_2021_em_status_occs_Flanders <dbl>,
## # year_2020_em_status_occupancy_Flanders <dbl>,
## # year_2020_em_status_occs_Flanders <dbl>, mean_growth <dbl>, …
##
## $Wallonia
## # A tibble: 866 × 13
## taxonKey canonicalName kingdom class year_2022_em_status_…¹
## <dbl> <chr> <chr> <chr> <dbl>
## 1 3033242 Anemone blanda Plantae Magnoliops… 3
## 2 3098912 Cosmos bipinnatus Plantae Magnoliops… 3
## 3 3152583 Hibiscus syriacus Plantae Magnoliops… 3
## 4 3012089 Sorbaria sorbifolia Plantae Magnoliops… 3
## 5 2888439 Papaver somniferum Plantae Magnoliops… 3
## 6 5362054 Crassula helmsii Plantae Magnoliops… 3
## 7 3084022 Phytolacca acinosa Plantae Magnoliops… 3
## 8 5281901 Campylopus introflexus Plantae Bryopsida 3
## 9 2226990 Pacifastacus leniusculus Animalia Malacostra… 3
## 10 5285637 Pinus sylvestris Plantae Pinopsida 3
## # ℹ 856 more rows
## # ℹ abbreviated name: ¹year_2022_em_status_occupancy_Wallonia
## # ℹ 8 more variables: year_2022_em_status_occs_Wallonia <dbl>,
## # year_2021_em_status_occupancy_Wallonia <dbl>,
## # year_2021_em_status_occs_Wallonia <dbl>,
## # year_2020_em_status_occupancy_Wallonia <dbl>,
## # year_2020_em_status_occs_Wallonia <dbl>, mean_growth <dbl>, …
##
## $Brussels
## # A tibble: 505 × 13
## taxonKey canonicalName kingdom class year_2022_em_status_o…¹
## <dbl> <chr> <chr> <chr> <dbl>
## 1 2037925 Orientus ishidae Animalia Insecta 3
## 2 1311477 Vespa velutina Animalia Insecta 3
## 3 2153287 Zoropsis spinimana Animalia Arachnida 3
## 4 3012089 Sorbaria sorbifolia Plantae Magnoliopsida 3
## 5 3054117 Aubrieta deltoidea Plantae Magnoliopsida 3
## 6 7127810 Cyclamen hederifolium Plantae Magnoliopsida 3
## 7 8313153 Quercus palustris Plantae Magnoliopsida 3
## 8 3067119 Euphorbia maculata Plantae Magnoliopsida 3
## 9 2891783 Impatiens balfourii Plantae Magnoliopsida 3
## 10 3084015 Phytolacca americana Plantae Magnoliopsida 3
## # ℹ 495 more rows
## # ℹ abbreviated name: ¹year_2022_em_status_occupancy_Brussels
## # ℹ 8 more variables: year_2022_em_status_occs_Brussels <dbl>,
## # year_2021_em_status_occupancy_Brussels <dbl>,
## # year_2021_em_status_occs_Brussels <dbl>,
## # year_2020_em_status_occupancy_Brussels <dbl>,
## # year_2020_em_status_occs_Brussels <dbl>, mean_growth <dbl>, …We can also rank using a point strategy reflecting the same ranking strategy as before. To do so we apply gain factors to all emerging status scores using the number of observations in 2020 in all Belgium/region as reference (gain factor = 1).
The following gain factors are applied for Belgium:
| indicator | year | region | gain factor |
|---|---|---|---|
| occupancy | 2022 | protected areas | 3 |
| observations | 2022 | protected areas | 2.5 |
| occupancy | 2022 | Belgium | 2.5 |
| observations | 2022 | Belgium | 2 |
| occupancy | 2021 | protected areas | 2.5 |
| observations | 2021 | protected areas | 2 |
| occupancy | 2021 | Belgium | 2 |
| observations | 2021 | Belgium | 1.5 |
| occupancy | 2020 | protected areas | 2 |
| observations | 2020 | protected areas | 1.5 |
| occupancy | 2020 | Belgium | 1.5 |
| observations | 2020 | Belgium | 1 |
For regions, the gain factors are the same, but the ones for protected areas are not applied.
The following gain factors are applied for each region:
| indicator | year | gain factor |
|---|---|---|
| occupancy | 2022 | 2.5 |
| observations | 2022 | 2 |
| occupancy | 2021 | 2 |
| observations | 2021 | 1.5 |
| occupancy | 2020 | 1.5 |
| observations | 2020 | 1 |
ranking_pts_be <-
em_df_be %>%
tidylog::mutate(em_pts = year_2022_em_status_occupancy_natura2000 * 3 +
year_2022_em_status_occs_natura2000 * 2.5 +
year_2022_em_status_occupancy_Belgium * 2.5 +
year_2022_em_status_occs_Belgium * 2 +
year_2021_em_status_occupancy_natura2000 * 2.5 +
year_2021_em_status_occs_natura2000 * 2 +
year_2021_em_status_occupancy_Belgium * 2 +
year_2021_em_status_occs_Belgium * 1.5 +
year_2020_em_status_occupancy_natura2000 * 2 +
year_2020_em_status_occs_natura2000 * 1.5 +
year_2020_em_status_occupancy_Belgium * 1.5 +
year_2020_em_status_occs_Belgium) %>%
dplyr::arrange(
desc(em_pts),
desc(mean_growth)
) %>%
tidylog::select(
taxonKey, canonicalName, kingdom, class,
em_pts, mean_growth,
year_2022_em_status_occupancy_natura2000,
year_2022_em_status_occs_natura2000,
year_2022_em_status_occupancy_Belgium,
year_2022_em_status_occs_Belgium,
year_2021_em_status_occupancy_natura2000,
year_2021_em_status_occs_natura2000,
year_2021_em_status_occupancy_Belgium,
year_2021_em_status_occs_Belgium,
year_2020_em_status_occupancy_natura2000,
year_2020_em_status_occs_natura2000,
year_2020_em_status_occupancy_Belgium,
year_2020_em_status_occs_Belgium,
dplyr::everything()
)
ranking_pts_beranking_pts_regions <- purrr::imap(
em_df_regions,
function(em_df, region) {
em_df %>%
# Remove the region from the column names for generalization
tidylog::rename_with(~ gsub(paste0("_", region, "$"), "", .x)) %>%
tidylog::mutate(em_pts = year_2022_em_status_occupancy * 2.5 +
year_2022_em_status_occs * 2 +
year_2021_em_status_occupancy * 2 +
year_2021_em_status_occs * 1.5 +
year_2020_em_status_occupancy * 1.5 +
year_2020_em_status_occs) %>%
dplyr::arrange(
desc(em_pts),
desc(mean_growth)
) %>%
tidylog::select(
taxonKey, canonicalName, kingdom, class,
em_pts, mean_growth,
year_2022_em_status_occupancy,
year_2022_em_status_occs,
year_2021_em_status_occupancy,
year_2021_em_status_occs,
year_2020_em_status_occupancy,
year_2020_em_status_occs,
dplyr::everything()
) %>%
# Add region back to the column names starting with `year`
tidylog::rename_with(~ paste0(.x, "_", region), starts_with("year"))
}
)
ranking_pts_regions## $Flanders
## # A tibble: 1,655 × 14
## taxonKey canonicalName kingdom class em_pts mean_growth
## <dbl> <chr> <chr> <chr> <dbl> <dbl>
## 1 6098843 Ctenolepisma longicaudatum Animalia Insecta 31.5 1.95
## 2 2225772 Hemigrapsus sanguineus Animalia Malacostraca 31.5 1.52
## 3 1047536 Leptinotarsa decemlineata Animalia Insecta 31.5 1.39
## 4 2498344 Cygnus atratus Animalia Aves 31.5 1.27
## 5 3098912 Cosmos bipinnatus Plantae Magnoliopsida 31.5 1.23
## 6 5232437 Branta canadensis Animalia Aves 31.5 1.22
## 7 2479226 Psittacula krameri Animalia Aves 31.5 1.21
## 8 2495494 Geopelia cuneata Animalia Aves 31.5 1.20
## 9 2650827 Cyrtomium fortunei Plantae Polypodiopsi… 31.5 1.18
## 10 6096602 Lophocolea semiteres Plantae Jungermannio… 31.5 1.17
## # ℹ 1,645 more rows
## # ℹ 8 more variables: year_2022_em_status_occupancy_Flanders <dbl>,
## # year_2022_em_status_occs_Flanders <dbl>,
## # year_2021_em_status_occupancy_Flanders <dbl>,
## # year_2021_em_status_occs_Flanders <dbl>,
## # year_2020_em_status_occupancy_Flanders <dbl>,
## # year_2020_em_status_occs_Flanders <dbl>, kingdomKey <dbl>, classKey <dbl>
##
## $Wallonia
## # A tibble: 866 × 14
## taxonKey canonicalName kingdom class em_pts mean_growth
## <dbl> <chr> <chr> <chr> <dbl> <dbl>
## 1 3033242 Anemone blanda Plantae Magnoliopsida 31.5 1.47
## 2 3098912 Cosmos bipinnatus Plantae Magnoliopsida 31.5 1.46
## 3 3152583 Hibiscus syriacus Plantae Magnoliopsida 31.5 1.41
## 4 3012089 Sorbaria sorbifolia Plantae Magnoliopsida 31.5 1.35
## 5 2888439 Papaver somniferum Plantae Magnoliopsida 31.5 1.19
## 6 5362054 Crassula helmsii Plantae Magnoliopsida 31.5 1.16
## 7 3084022 Phytolacca acinosa Plantae Magnoliopsida 31.5 1.13
## 8 5281901 Campylopus introflexus Plantae Bryopsida 31.5 1.07
## 9 2226990 Pacifastacus leniusculus Animalia Malacostraca 31.5 1.03
## 10 5285637 Pinus sylvestris Plantae Pinopsida 31.5 1.02
## # ℹ 856 more rows
## # ℹ 8 more variables: year_2022_em_status_occupancy_Wallonia <dbl>,
## # year_2022_em_status_occs_Wallonia <dbl>,
## # year_2021_em_status_occupancy_Wallonia <dbl>,
## # year_2021_em_status_occs_Wallonia <dbl>,
## # year_2020_em_status_occupancy_Wallonia <dbl>,
## # year_2020_em_status_occs_Wallonia <dbl>, kingdomKey <dbl>, classKey <dbl>
##
## $Brussels
## # A tibble: 505 × 14
## taxonKey canonicalName kingdom class em_pts mean_growth
## <dbl> <chr> <chr> <chr> <dbl> <dbl>
## 1 2037925 Orientus ishidae Animalia Insecta 31.5 1.16
## 2 1311477 Vespa velutina Animalia Insecta 31.5 NaN
## 3 2153287 Zoropsis spinimana Animalia Arachnida 31.5 NaN
## 4 3012089 Sorbaria sorbifolia Plantae Magnoliopsida 31.5 NaN
## 5 3054117 Aubrieta deltoidea Plantae Magnoliopsida 31.5 NaN
## 6 7127810 Cyclamen hederifolium Plantae Magnoliopsida 31.5 NaN
## 7 8313153 Quercus palustris Plantae Magnoliopsida 31.5 NaN
## 8 2164198 Cheiracanthium mildei Animalia Arachnida 30 NaN
## 9 3147168 Erigeron karvinskianus Plantae Magnoliopsida 30 NaN
## 10 5341201 Anchusa officinalis Plantae Magnoliopsida 29.5 NaN
## # ℹ 495 more rows
## # ℹ 8 more variables: year_2022_em_status_occupancy_Brussels <dbl>,
## # year_2022_em_status_occs_Brussels <dbl>,
## # year_2021_em_status_occupancy_Brussels <dbl>,
## # year_2021_em_status_occs_Brussels <dbl>,
## # year_2020_em_status_occupancy_Brussels <dbl>,
## # year_2020_em_status_occs_Brussels <dbl>, kingdomKey <dbl>, classKey <dbl>We save the final emerging status for Belgium,
em_dfbe:
readr::write_tsv(em_df_be,
file = here::here("data", "output", "emerging_status_Belgium.tsv"),
na = ""
)and the final emerging status for regions,
em_df_regions:
purrr::iwalk(
em_df_regions,
function(df, region) {
readr::write_tsv(df,
file = here::here("data", "output", paste0("emerging_status_", region, ".tsv")),
na = ""
)
}
)We save both the hierarchical ranking and the ranking based on points strategy for Belgium and its regions.
The ranking for Belgium based on hierarchical strategy is saved in
data/output/ranking_emerging_status_hierarchical_strategy_Belgium.tsv:
readr::write_tsv(ranking_be,
file = here::here(
"data",
"output",
"ranking_emerging_status_hierarchical_strategy_Belgium.tsv"
),
na = ""
)The ranking for the regions based on hierarchical strategy are saved
in
data/output/ranking_emerging_status_hierarchical_strategy_*.tsv
where * is the region:
purrr::iwalk(
ranking_regions,
function(df, region) {
readr::write_tsv(df,
file = here::here(
"data",
"output",
paste0("ranking_emerging_status_hierarchical_strategy_", region, ".tsv")
),
na = ""
)
}
)Ranking for Belgium based on points strategy is saved in
data/output/ranking_emerging_status_points_strategy_Belgium.tsv:
readr::write_tsv(ranking_pts_be,
file = here::here(
"data",
"output",
"ranking_emerging_status_points_strategy_Belgium.tsv"
),
na = ""
)Ranking for the regions based on points strategy are saved in
data/output/ranking_emerging_status_points_strategy_*.tsv
where * is the region:
purrr::iwalk(
ranking_pts_regions,
function(df, region) {
readr::write_tsv(df,
file = here::here(
"data",
"output",
paste0("ranking_emerging_status_points_strategy_", region, ".tsv")
),
na = ""
)
}
)