Introduction to danstat

library(danstat)
library(purrr)
library(dplyr)
#> 
#> Attaching package: 'dplyr'
#> The following objects are masked from 'package:stats':
#> 
#>     filter, lag
#> The following objects are masked from 'package:base':
#> 
#>     intersect, setdiff, setequal, union
library(ggplot2)
library(kableExtra)
#> 
#> Attaching package: 'kableExtra'
#> The following object is masked from 'package:dplyr':
#> 
#>     group_rows

get_subjects()
#>    id            description active hasSubjects subjects
#> 1   1                 People   TRUE        TRUE     NULL
#> 2   2      Labour and income   TRUE        TRUE     NULL
#> 3   3                Economy   TRUE        TRUE     NULL
#> 4   4      Social conditions   TRUE        TRUE     NULL
#> 5   5 Education and research   TRUE        TRUE     NULL
#> 6   6               Business   TRUE        TRUE     NULL
#> 7   7              Transport   TRUE        TRUE     NULL
#> 8   8    Culture and leisure   TRUE        TRUE     NULL
#> 9   9 Environment and energy   TRUE        TRUE     NULL
#> 10 19                  Other   TRUE        TRUE     NULL

subj <- get_subjects(subjects = c("6","7"))
subsubjects <- subj$subjects %>% bind_rows()
subsubjects
#>      id                            description active hasSubjects subjects
#> 1  3456                  Structure of business   TRUE        TRUE     NULL
#> 2  3457                       Business economy   TRUE        TRUE     NULL
#> 3  3458              International enterprises   TRUE        TRUE     NULL
#> 4  3460 Agriculture, horticulture and forestry   TRUE        TRUE     NULL
#> 5  3461                Fishery and aquaculture   TRUE        TRUE     NULL
#> 6  3462               Manufacturing industries   TRUE        TRUE     NULL
#> 7  3463                           Construction   TRUE        TRUE     NULL
#> 8  3465                                  Trade   TRUE        TRUE     NULL
#> 9  3466               Accommodation and travel   TRUE        TRUE     NULL
#> 10 3450                       Financial sector   TRUE        TRUE     NULL
#> 11 3467                         Service sector   TRUE        TRUE     NULL
#> 12 3459              Business tendency surveys   TRUE        TRUE     NULL
#> 13 3440                     Means of transport   TRUE        TRUE     NULL
#> 14 3469             Traffic and infrastructure   TRUE        TRUE     NULL
#> 15 3115                    Passenger transport   TRUE        TRUE     NULL
#> 16 3464                        Goods transport   TRUE        TRUE     NULL
#> 17 3413                      Traffic accidents   TRUE        TRUE     NULL

tables <- get_tables(subjects = c("3465", "3413")) 
tables %>% 
  select(id, text, variables) %>% 
  kable()

vars_acc <- get_table_metadata(table_id = "uheld4", variables_only = TRUE)
vars_alco <- get_table_metadata(table_id = "alko3", variables_only = TRUE)

vars_acc %>% 
  select(id, text)
#>         id               text
#> 1   UHELDA   type of accident
#> 2 TRANSMID means of transport
#> 3     KLOK               hour
#> 4      UGE    day of the week
#> 5      MND              month
#> 6      Tid               time

vars_alco %>% 
  select(id, text)
#>     id text
#> 1 TYPE type
#> 2  Tid time

vars_acc$values[1] %>% 
  kable()

vars_alco$values[1] %>% 
  kable()

variable_codes <- vars_acc$id[c(1, 3, 6)] # UHELDA, KLOK and Tid
variable_values <- list(c(1000, 2000), NA, NA) # all values for KLOK and Tid

# Construct the variable_input as a list of code-values pairs
variable_input <- purrr::map2(.x = variable_codes, .y = variable_values, .f = ~list(code = .x, values = .y))

# Get data 
accidents <- get_data("uheld4", variables = variable_input)
#> Rows: 1350 Columns: 4
#> ── Column specification ────────────────────────────────────────────────────────
#> Delimiter: ";"
#> chr (2): UHELDA, KLOK
#> dbl (2): TID, INDHOLD
#> 
#> ℹ Use `spec()` to retrieve the full column specification for this data.
#> ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.
head(accidents) %>% kable()

variable_codes <- vars_alco$id 
variable_values <- list(c("055", "09"), NA) # All values for Tid

# Construct the variable_input as a list of code-values pairs
variable_input <- purrr::map2(.x = variable_codes, .y = variable_values, .f = ~list(code = .x, values = .y))

# Get data 
alcohol <- get_data("alko3", variables = variable_input)
#> Rows: 132 Columns: 3
#> ── Column specification ────────────────────────────────────────────────────────
#> Delimiter: ";"
#> chr (2): TYPE, INDHOLD
#> dbl (1): TID
#> 
#> ℹ Use `spec()` to retrieve the full column specification for this data.
#> ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.
alcohol %>% 
  filter(INDHOLD != "..") %>% # the API returns ".." as missing values
  head() %>% 
  kable()

accidents_by_hour <- accidents %>%
    filter(KLOK != "Not stated") %>%
    group_by(UHELDA, KLOK) %>%
    summarise(mean_accidents = mean(INDHOLD, na.rm = TRUE))
#> `summarise()` has grouped output by 'UHELDA'. You can override using the
#> `.groups` argument.

accidents_by_hour %>%
    ggplot(aes(x = KLOK, y = mean_accidents, color = UHELDA, group = UHELDA)) +
    geom_line() +
    geom_point() +
    theme_bw() + 
  theme(legend.position="top") +
  labs(x = "Time of day", y = "Average annual accidents")

accidents_by_year <- accidents %>% 
  group_by(UHELDA, TID) %>% 
  summarize(INDHOLD = sum(INDHOLD)) %>% 
  ungroup()
#> `summarise()` has grouped output by 'UHELDA'. You can override using the
#> `.groups` argument.

accidents_by_year %>%
  group_by(UHELDA) %>% 
  summarise(min(TID), max(TID))
#> # A tibble: 2 × 3
#>   UHELDA                                            `min(TID)` `max(TID)`
#>   <chr>                                                  <dbl>      <dbl>
#> 1 Accidents involving persons influenced by alcohol       1997       2023
#> 2 Others accidents                                        1997       2023

alcohol_by_year <- alcohol %>% 
  filter(INDHOLD != "..") %>% # the API returns ".." as missing values
  mutate(INDHOLD = as.numeric(INDHOLD))

alcohol_by_year %>%
  group_by(TYPE) %>% 
  summarise(min(TID), max(TID))
#> # A tibble: 2 × 3
#>   TYPE                                                     `min(TID)` `max(TID)`
#>   <chr>                                                         <dbl>      <dbl>
#> 1 Consumption of lager equivalents, total, excl. tax free…       2000       2010
#> 2 Sales of lager equivalents, total excl. tax free beer (…       1955       2020

alcohol_data <- alcohol_by_year %>% 
  filter(between(TID, 1997, 2021),
         grepl("sales", TYPE, ignore.case = TRUE)) %>% 
  select(year = TID,
         alcohol_sales = INDHOLD)

accidents_data <- accidents_by_year %>% 
  filter(grepl("alcohol", UHELDA, ignore.case = TRUE)) %>% 
  select(year = TID,
         alcohol_accidents = INDHOLD)

accidents_2000 <- accidents_data %>% 
  filter(year == 2000) %>% 
  pull(alcohol_accidents)
  
accidents_data$alcohol_accidents <- round((accidents_data$alcohol_accidents/accidents_2000)*100)

data <- inner_join(alcohol_data, accidents_data, by = "year") %>% 
  mutate(risk_index = round(alcohol_accidents/alcohol_sales*100))

data %>% 
  ggplot(aes(x=year, y=risk_index)) +
  geom_col()