1. Packages
need <- c(
"sf","dplyr","stringr","readr","ggplot2","tmap","spdep",
"flextable","officer","glue","tidyr"
)
inst <- need[!need %in% installed.packages()[, "Package"]]
if (length(inst)) install.packages(inst, dependencies = TRUE)
library(sf)
library(dplyr)
library(stringr)
library(readr)
library(ggplot2)
library(tmap)
library(spdep)
library(flextable)
library(officer)
library(glue)
library(tidyr)
2. Data & Basic
Checks
data_path <- params$data_path
out_dir <- params$out_dir
if (!dir.exists(out_dir)) dir.create(out_dir, recursive = TRUE)
dist_data_shp <- readRDS(data_path)
# Peek
class(dist_data_shp)
## [1] "sf" "tbl_df" "tbl" "data.frame"
names(dist_data_shp)[1:20]
## [1] "GID_2" "GID_0" "COUNTRY"
## [4] "GID_1" "NAME_1" "NL_NAME_1"
## [7] "NAME_2" "VARNAME_2" "NL_NAME_2"
## [10] "TYPE_2" "ENGTYPE_2" "CC_2"
## [13] "HASC_2" "geometry" "district_id"
## [16] "Year" "Tyhoid_cases" "Population"
## [19] "Typhoid_Prevalence" "incidence_rate"
head(dist_data_shp)
# Geometry type
sf::st_geometry_type(dist_data_shp)[1:5]
## [1] MULTIPOLYGON MULTIPOLYGON MULTIPOLYGON MULTIPOLYGON MULTIPOLYGON
## 18 Levels: GEOMETRY POINT LINESTRING POLYGON MULTIPOINT ... TRIANGLE
# Standardize needed columns
dat <- dist_data_shp %>%
transmute(
district_id = district_id,
region = NAME_1,
district = NAME_2,
year = as.integer(Year),
cases = as.numeric(`Tyhoid_cases`),
pop = as.numeric(Population),
geometry
) %>%
filter(!is.na(year), year >= 2015, year <= 2024) %>%
filter(!is.na(cases), !is.na(pop), cases >= 0, pop > 0) %>%
mutate(
district_std = str_squish(str_to_lower(gsub("[^A-Za-z0-9 ]", " ", district))),
region_std = str_squish(str_to_lower(gsub("[^A-Za-z0-9 ]", " ", region)))
)
# Summary
dat %>% summarise(
n_rows = n(),
n_years = dplyr::n_distinct(year),
years = paste0(min(year), "–", max(year)),
n_districts = dplyr::n_distinct(district_id),
n_regions = dplyr::n_distinct(region)
)
3. National Rate by
Year & District-Year SIR
nat_by_year <- dat %>%
sf::st_drop_geometry() %>%
group_by(year) %>%
summarise(
nat_cases = sum(cases, na.rm = TRUE),
nat_pop = sum(pop, na.rm = TRUE),
nat_rate = nat_cases / nat_pop,
.groups = "drop"
)
sir_yearly <- dat %>%
left_join(nat_by_year, by = "year") %>%
mutate(
E = nat_rate * pop,
SIR = ifelse(E > 0, cases / E, NA_real_),
SIR_low = ifelse(cases == 0, 0, qchisq(0.025, 2*cases) / (2*E)),
SIR_high = qchisq(0.975, 2*(cases + 1)) / (2*E),
sig_flag = dplyr::case_when(
SIR_low > 1 ~ "High (significant)",
SIR_high < 1 ~ "Low (significant)",
TRUE ~ "Not significant"
)
)
# Check observed ≈ expected nationally per year
sir_yearly %>%
sf::st_drop_geometry() %>%
group_by(year) %>%
summarise(obs = sum(cases), Exp = sum(E)) %>%
mutate(obs_to_Exp = obs/Exp)
4. Overall (2015–2024)
District SIR
sir_overall <- sir_yearly %>%
sf::st_drop_geometry() %>%
group_by(district_id, region, district, district_std, region_std) %>%
summarise(
obs = sum(cases, na.rm = TRUE),
Exp = sum(E, na.rm = TRUE),
.groups = "drop"
) %>%
mutate(
SIR = ifelse(Exp > 0, obs / Exp, NA_real_),
SIR_low = ifelse(obs == 0, 0, qchisq(0.025, 2*obs) / (2*Exp)),
SIR_high = qchisq(0.975, 2*(obs + 1)) / (2*Exp),
sig_flag = dplyr::case_when(
SIR_low > 1 ~ "High (significant)",
SIR_high < 1 ~ "Low (significant)",
TRUE ~ "Not significant"
)
) %>%
arrange(desc(SIR))
summary(sir_overall$SIR)
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 0.05205 0.43194 0.79123 0.99661 1.25579 10.07133
table(sir_overall$sig_flag, useNA = "ifany")
##
## High (significant) Low (significant) Not significant
## 89 169 2
5. Tables: Top/Bottom
10 (Baseline & Stable)
# Baseline top/bottom 10
top10 <- sir_overall %>% filter(!is.na(SIR)) %>% arrange(desc(SIR)) %>% slice_head(n = 10) %>% mutate(Rank = row_number())
bottom10 <- sir_overall %>% filter(!is.na(SIR)) %>% arrange(SIR) %>% slice_head(n = 10) %>% mutate(Rank = row_number())
fmt_table <- function(df) {
df %>%
mutate(
Exp = round(Exp, 1),
SIR = round(SIR, 2),
SIR_low = round(SIR_low, 2),
SIR_high = round(SIR_high, 2),
`95% CI` = paste0(SIR_low, "–", SIR_high)
) %>%
select(Rank, Region = region, District = district,
Observed = obs, Expected = Exp, SIR, `95% CI`, Significance = sig_flag)
}
top10_tbl <- fmt_table(top10)
bottom10_tbl <- fmt_table(bottom10)
# Stability filter (Exp >= 20)
stable_overall <- sir_overall %>% filter(Exp >= 20)
top10_stable <- stable_overall %>% arrange(desc(SIR)) %>% slice_head(n = 10) %>% mutate(Rank = row_number())
bottom10_stable <- stable_overall %>% arrange(SIR) %>% slice_head(n = 10) %>% mutate(Rank = row_number())
top10_stable_tbl <- fmt_table(top10_stable)
bottom10_stable_tbl <- fmt_table(bottom10_stable)
# Export CSVs
readr::write_csv(sir_yearly %>% sf::st_drop_geometry() %>%
select(district_id, region, district, year, cases, pop, E, SIR, SIR_low, SIR_high, sig_flag),
file.path(out_dir, "district_year_SIR_2015_2024.csv"))
readr::write_csv(sir_overall %>% select(district_id, region, district, obs, Exp, SIR, SIR_low, SIR_high, sig_flag),
file.path(out_dir, "district_overall_SIR_2015_2024.csv"))
readr::write_csv(top10_tbl, file.path(out_dir, "Top10_Districts_SIR_2015_2024.csv"))
readr::write_csv(bottom10_tbl, file.path(out_dir, "Bottom10_Districts_SIR_2015_2024.csv"))
readr::write_csv(top10_stable_tbl, file.path(out_dir, "Top10_Districts_SIR_Stable_Exp20.csv"))
readr::write_csv(bottom10_stable_tbl, file.path(out_dir, "Bottom10_Districts_SIR_Stable_Exp20.csv"))
doc <- read_docx()
doc <- body_add_par(doc, "Top 10 Districts by Overall SIR (2015–2024)", style = "heading 2")
doc <- body_add_flextable(doc, flextable(top10_tbl) %>% autofit())
doc <- body_add_par(doc, "Bottom 10 Districts by Overall SIR (2015–2024)", style = "heading 2")
doc <- body_add_flextable(doc, flextable(bottom10_tbl) %>% autofit())
doc <- body_add_par(doc, "Top 10 Districts by Overall SIR (Stable: Expected ≥ 20)", style = "heading 2")
doc <- body_add_flextable(doc, flextable(top10_stable_tbl) %>% autofit())
doc <- body_add_par(doc, "Bottom 10 Districts by Overall SIR (Stable: Expected ≥ 20)", style = "heading 2")
doc <- body_add_flextable(doc, flextable(bottom10_stable_tbl) %>% autofit())
print(doc, target = file.path(out_dir, "District_SIR_TopBottom10.docx"))
6. Maps — Overall SIR
& Significant High-Risk Outlines
# Build sf for mapping
dist_geom <- dat %>%
select(district_id, region, district, district_std, region_std, geometry) %>%
distinct(district_id, .keep_all = TRUE)
sir_overall_sf <- dist_geom %>%
left_join(sir_overall, by = c("district_id","region","district","district_std","region_std")) %>%
st_make_valid()
# ggplot map
p_sir <- ggplot(sir_overall_sf) +
geom_sf(aes(fill = SIR), linewidth = 0.1, color = "grey30") +
scale_fill_gradient(trans = "sqrt", name = "SIR") +
labs(title = "Overall District SIR for Typhoid Fever (2015–2024)",
caption = "Source: DHIMS II") +
theme_minimal(base_size = 11) +
theme(panel.grid.major = element_blank(), axis.text = element_blank(), axis.title = element_blank())
p_sir
ggsave(file.path(out_dir, "Map1_overall_SIR_2015_2024.png"), p_sir, width = 10, height = 8, dpi = 300)
# tmap (plot mode for saving)
tmap_mode("plot")
m_sir <- tm_shape(sir_overall_sf) +
tm_polygons("SIR", style = "quantile", n = 5, palette = "Oranges", title = "SIR (overall)") +
tm_layout(title = "Overall District SIR (2015–2024)", legend.outside = TRUE, frame = FALSE)
m_sir
tmap_save(m_sir, file.path(out_dir, "Map2_overall_SIR_tmap_2015_2024.png"), width = 10, height = 8, dpi = 300)
# Significant high-risk outlines
sig_sf <- sir_overall_sf %>%
mutate(sig_simple = dplyr::case_when(
SIR_low > 1 ~ "High (sig.)",
SIR_high < 1 ~ "Low (sig.)",
TRUE ~ "NS"
))
p_sig <- ggplot() +
geom_sf(data = sig_sf, aes(fill = SIR), linewidth = 0.1, color = "grey30") +
geom_sf(data = subset(sig_sf, sig_simple == "High (sig.)"),
fill = NA, linewidth = 0.8, linetype = 2, color = "black") +
scale_fill_gradient(trans = "sqrt", name = "SIR") +
labs(title = "District SIR with Significant High-Risk Outlines (2015–2024)",
subtitle = "Dashed outline = SIR 95% CI lower bound > 1",
caption = "Source: DHIMS II") +
theme_minimal(base_size = 11) +
theme(panel.grid.major = element_blank(), axis.text = element_blank(), axis.title = element_blank())
p_sig
ggsave(file.path(out_dir, "Map3_overall_SIR_sig_highlight.png"), p_sig, width = 10, height = 8, dpi = 300)
7. Faceted
District-Year SIR Map (Appendix)
sir_yearly_plot <- sir_yearly %>%
filter(!is.na(SIR), is.finite(SIR), year >= 2015, year <= 2024)
upper_cap <- quantile(sir_yearly_plot$SIR, 0.95, na.rm = TRUE)
sir_yearly_plot <- sir_yearly_plot %>% mutate(SIR_capped = pmin(SIR, upper_cap))
p_faceted_cont <- ggplot(sir_yearly_plot) +
geom_sf(aes(fill = SIR_capped), linewidth = 0.08, color = "grey30") +
scale_fill_gradient(low = "#FFF0B3", high = "#E65100", trans = "sqrt",
limits = c(0, upper_cap), breaks = pretty(c(0, upper_cap), 5), name = "SIR") +
facet_wrap(~ year, ncol = 5) +
labs(title = "District SIR by Year (2015–2024)",
subtitle = "SIR = Observed / Expected (expected uses national annual rate)",
caption = "Source: DHIMS II") +
theme_minimal(base_size = 10) +
theme(panel.grid.major = element_blank(), axis.text = element_blank(),
axis.title = element_blank(), strip.text = element_text(face = "bold"))
p_faceted_cont
ggsave(file.path(out_dir, "Appendix_FacetMap_District_SIR_Continuous_2015_2024.png"),
p_faceted_cont, width = 14, height = 10, dpi = 300)
ggsave(file.path(out_dir, "Appendix_FacetMap_District_SIR_Continuous_2015_2024.pdf"),
p_faceted_cont, width = 14, height = 10)
8. Spatial
Autocorrelation & Clusters (Moran’s I, LISA)
# Queen contiguity
nb <- poly2nb(sir_overall_sf, queen = TRUE)
lw <- nb2listw(nb, style = "W", zero.policy = TRUE)
# Global Moran's I (greater = positive autocorrelation)
x <- log(sir_overall_sf$SIR)
m_global <- moran.test(x, lw, alternative = "greater", zero.policy = TRUE)
m_perm <- moran.mc(x, lw, nsim = 9999, alternative = "greater", zero.policy = TRUE)
m_global
##
## Moran I test under randomisation
##
## data: x
## weights: lw
##
## Moran I statistic standard deviate = 3.27, p-value = 0.0005378
## alternative hypothesis: greater
## sample estimates:
## Moran I statistic Expectation Variance
## 0.121037898 -0.003861004 0.001458897
m_perm
##
## Monte-Carlo simulation of Moran I
##
## data: x
## weights: lw
## number of simulations + 1: 10000
##
## statistic = 0.12104, observed rank = 9992, p-value = 8e-04
## alternative hypothesis: greater
# Local Moran's I (LISA)
x_raw <- sir_overall_sf$SIR; x_raw[!is.finite(x_raw)] <- NA_real_
z <- as.numeric(scale(log1p(x_raw))) # standardized numeric vector
lisa <- localmoran(z, lw, zero.policy = TRUE, na.action = na.exclude)
Ii <- lisa[, "Ii"]
Z <- lisa[, "Z.Ii"]
p_raw <- 2 * pnorm(-abs(Z))
p_fdr <- p.adjust(p_raw, method = "fdr")
sir_overall_sf$Ii <- Ii
sir_overall_sf$Z.Ii <- Z
sir_overall_sf$P.FDR <- p_fdr
ok <- is.finite(Ii) & is.finite(z) & is.finite(p_fdr)
sir_overall_sf$LISA_cat <- "NS"
sir_overall_sf$LISA_cat[ok & z > 0 & Ii > 0 & p_fdr < 0.05] <- "High-High"
sir_overall_sf$LISA_cat[ok & z < 0 & Ii > 0 & p_fdr < 0.05] <- "Low-Low"
sir_overall_sf$LISA_cat[ok & z > 0 & Ii < 0 & p_fdr < 0.05] <- "High-Low"
sir_overall_sf$LISA_cat[ok & z < 0 & Ii < 0 & p_fdr < 0.05] <- "Low-High"
p_lisa <- ggplot(sir_overall_sf) +
geom_sf(aes(fill = LISA_cat), color = "grey30", linewidth = 0.1) +
scale_fill_manual(values = c("High-High"="#E65100","Low-Low"="#2B8CBE",
"High-Low"="#FDB863","Low-High"="#67A9CF","NS"="#F0F0F0"),
name = "LISA clusters") +
labs(title = "Local Moran’s I Clusters (Overall SIR, 2015–2024)",
subtitle = "Two-sided test; FDR-adjusted p < 0.05; variable = log1p(SIR), standardized",
caption = "Source: DHIMS II") +
theme_minimal() +
theme(axis.text = element_blank(), axis.title = element_blank())
p_lisa
ggsave(file.path(out_dir, "Map_LISA_Overall_2015_2024.png"), p_lisa, width = 10, height = 8, dpi = 300)
9. Getis–Ord Gi*
Hotspots
gi <- localG(z, lw, zero.policy = TRUE)
gi_z <- as.numeric(gi)
gi_p <- 2 * pnorm(-abs(gi_z))
gi_fdr <- p.adjust(gi_p, method = "fdr")
sir_overall_sf$GiZ <- gi_z
sir_overall_sf$GiFDR <- gi_fdr
sir_overall_sf$GiCat <- "NS"
sir_overall_sf$GiCat[gi_z > 0 & gi_fdr < 0.05] <- "Hotspot"
sir_overall_sf$GiCat[gi_z < 0 & gi_fdr < 0.05] <- "Coldspot"
p_gi <- ggplot(sir_overall_sf) +
geom_sf(aes(fill = GiCat), color = "grey30", linewidth = 0.1) +
scale_fill_manual(values = c("Hotspot"="#E65100","Coldspot"="#2B8CBE","NS"="#F0F0F0"),
name = "Getis–Ord Gi*") +
labs(title = "Getis–Ord Gi* Hotspots (Overall SIR, 2015–2024)",
subtitle = "Two-sided FDR-adjusted p < 0.05; variable = log1p(SIR), standardized",
caption = "Source: DHIMS II") +
theme_minimal() +
theme(axis.text = element_blank(), axis.title = element_blank())
p_gi
ggsave(file.path(out_dir, "Map_GiStar_Overall_2015_2024.png"), p_gi, width = 10, height = 8, dpi = 300)
# Counts to report
table(sir_overall_sf$GiCat)
##
## Hotspot NS
## 5 255
10. (Appendix)
Year-by-Year Gi* Hotspots Facet
dist_geom_key <- dat %>% select(district_id, geometry) %>% distinct(district_id, .keep_all = TRUE)
years <- sort(unique(sir_yearly$year))
gi_year_list <- vector("list", length(years)); names(gi_year_list) <- years
for (i in seq_along(years)) {
yy <- years[i]
d <- sir_yearly %>% filter(year == yy) %>% select(district_id, SIR) %>% st_drop_geometry()
d <- d[match(dist_geom_key$district_id, d$district_id), ]
d_sf <- dist_geom_key
d_sf$SIR <- d$SIR
xr <- d_sf$SIR; xr[!is.finite(xr)] <- NA_real_
zz <- as.numeric(scale(log1p(xr)))
giz <- as.numeric(localG(zz, lw, zero.policy = TRUE))
gip <- 2 * pnorm(-abs(giz))
gifdr <- p.adjust(gip, "fdr")
d_sf$GiCat <- "NS"
d_sf$GiCat[giz > 0 & gifdr < 0.05] <- "Hotspot"
d_sf$GiCat[giz < 0 & gifdr < 0.05] <- "Coldspot"
d_sf$year <- yy
gi_year_list[[i]] <- d_sf
}
gi_year_sf <- do.call(rbind, gi_year_list)
p_gi_year <- ggplot(gi_year_sf) +
geom_sf(aes(fill = GiCat), color = "grey30", linewidth = 0.05) +
scale_fill_manual(values = c("Hotspot"="#E65100","Coldspot"="#2B8CBE","NS"="#F0F0F0"), name = "Gi*") +
facet_wrap(~ year, ncol = 5) +
labs(title = "Getis–Ord Gi* Hotspots by Year (2015–2024)",
subtitle = "Two-sided FDR-adjusted p < 0.05; variable = log1p(SIR), standardized",
caption = "Source: DHIMS II") +
theme_minimal(base_size = 10) +
theme(panel.grid.major = element_blank(), axis.text = element_blank(),
axis.title = element_blank(), strip.text = element_text(face = "bold"))
p_gi_year
ggsave(file.path(out_dir, "Appendix_FacetMap_GiStar_2015_2024.png"), p_gi_year, width = 14, height = 10, dpi = 300)
# Yearly hotspot counts table
gi_counts_year <- gi_year_sf %>%
st_drop_geometry() %>%
count(year, GiCat) %>%
tidyr::pivot_wider(names_from = GiCat, values_from = n, values_fill = 0) %>%
arrange(year)
readr::write_csv(gi_counts_year, file.path(out_dir, "GiStar_counts_by_year.csv"))
11. Auto-filled Result
Sentences
sir_sum <- summary(sir_overall$SIR)
n_high <- sum(sir_overall$SIR_low > 1, na.rm = TRUE)
n_low <- sum(sir_overall$SIR_high < 1, na.rm = TRUE)
line_sir <- glue(
"Across districts, the overall SIR ranged from {round(sir_sum['Min.'],2)} to {round(sir_sum['Max.'],2)} ",
"(median {round(sir_sum['Median'],2)}). ",
"{n_high} districts had significantly higher risk (SIR 95% CI lower bound > 1), ",
"and {n_low} districts had significantly lower risk (upper bound < 1)."
)
line_sir
## Across districts, the overall SIR ranged from 0.05 to 10.07 (median 0.79). 89 districts had significantly higher risk (SIR 95% CI lower bound > 1), and 169 districts had significantly lower risk (upper bound < 1).
# Global Moran's I summary (printed earlier too)
m_global
##
## Moran I test under randomisation
##
## data: x
## weights: lw
##
## Moran I statistic standard deviate = 3.27, p-value = 0.0005378
## alternative hypothesis: greater
## sample estimates:
## Moran I statistic Expectation Variance
## 0.121037898 -0.003861004 0.001458897
m_perm
##
## Monte-Carlo simulation of Moran I
##
## data: x
## weights: lw
## number of simulations + 1: 10000
##
## statistic = 0.12104, observed rank = 9992, p-value = 8e-04
## alternative hypothesis: greater
12. Session Info
sessionInfo()
## R version 4.3.3 (2024-02-29 ucrt)
## Platform: x86_64-w64-mingw32/x64 (64-bit)
## Running under: Windows 11 x64 (build 26100)
##
## Matrix products: default
##
##
## locale:
## [1] LC_COLLATE=English_United States.utf8
## [2] LC_CTYPE=English_United States.utf8
## [3] LC_MONETARY=English_United States.utf8
## [4] LC_NUMERIC=C
## [5] LC_TIME=English_United States.utf8
##
## time zone: Atlantic/Reykjavik
## tzcode source: internal
##
## attached base packages:
## [1] stats graphics grDevices utils datasets methods base
##
## other attached packages:
## [1] tidyr_1.3.1 glue_1.7.0 officer_0.6.6 flextable_0.9.6
## [5] spdep_1.3-5 spData_2.3.1 tmap_3.3-4 ggplot2_3.5.1
## [9] readr_2.1.5 stringr_1.5.1 dplyr_1.1.4 sf_1.0-16
##
## loaded via a namespace (and not attached):
## [1] DBI_1.2.3 deldir_2.0-4 tmaptools_3.1-1
## [4] s2_1.1.6 rlang_1.1.4 magrittr_2.0.3
## [7] e1071_1.7-14 compiler_4.3.3 png_0.1-8
## [10] systemfonts_1.1.0 vctrs_0.6.5 httpcode_0.3.0
## [13] pkgconfig_2.0.3 wk_0.9.2 crayon_1.5.3
## [16] fastmap_1.2.0 labeling_0.4.3 lwgeom_0.2-14
## [19] leafem_0.2.4 utf8_1.2.4 promises_1.3.0
## [22] rmarkdown_2.27 tzdb_0.4.0 ragg_1.3.2
## [25] bit_4.0.5 purrr_1.0.2 xfun_0.45
## [28] cachem_1.1.0 jsonlite_1.8.8 highr_0.11
## [31] later_1.3.2 uuid_1.2-0 terra_1.7-78
## [34] parallel_4.3.3 R6_2.5.1 bslib_0.7.0
## [37] stringi_1.8.4 RColorBrewer_1.1-3 boot_1.3-29
## [40] jquerylib_0.1.4 stars_0.6-5 Rcpp_1.0.12
## [43] knitr_1.48 base64enc_0.1-3 httpuv_1.6.15
## [46] tidyselect_1.2.1 rstudioapi_0.16.0 dichromat_2.0-0.1
## [49] abind_1.4-5 yaml_2.3.9 codetools_0.2-19
## [52] curl_5.2.1 lattice_0.22-5 tibble_3.2.1
## [55] leafsync_0.1.0 shiny_1.8.1.1 withr_3.0.0
## [58] askpass_1.2.0 evaluate_0.24.0 units_0.8-5
## [61] proxy_0.4-27 zip_2.3.1 xml2_1.3.6
## [64] pillar_1.9.0 KernSmooth_2.23-22 generics_0.1.3
## [67] vroom_1.6.5 sp_2.1-4 hms_1.1.3
## [70] munsell_0.5.1 scales_1.3.0 xtable_1.8-4
## [73] class_7.3-22 gdtools_0.3.7 tools_4.3.3
## [76] gfonts_0.2.0 data.table_1.15.4 XML_3.99-0.17
## [79] grid_4.3.3 crosstalk_1.2.1 colorspace_2.1-0
## [82] raster_3.6-26 cli_3.6.3 textshaping_0.4.0
## [85] fontBitstreamVera_0.1.1 fansi_1.0.6 viridisLite_0.4.2
## [88] gtable_0.3.6 sass_0.4.9 digest_0.6.36
## [91] fontquiver_0.2.1 classInt_0.4-10 crul_1.4.2
## [94] farver_2.1.2 htmlwidgets_1.6.4 htmltools_0.5.8.1
## [97] lifecycle_1.0.4 leaflet_2.2.2 mime_0.12
## [100] bit64_4.0.5 fontLiberation_0.1.0 openssl_2.2.0