-
Notifications
You must be signed in to change notification settings - Fork 0
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
- Loading branch information
1 parent
0b56954
commit e6ae0d2
Showing
4 changed files
with
216 additions
and
0 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,45 @@ | ||
kap_kod,kap_zkr,kap_name | ||
301,KPR,Kancelář prezidenta republiky | ||
302,K PSP,Poslanecká sněmovna Parlamentu | ||
303,K Sen,Senát Parlamentu | ||
304,ÚV,Úřad vlády ČR | ||
306,MZV,Ministerstvo zahraničních věcí | ||
307,MO,Ministerstvo obrany | ||
308,NBÚ,Národní bezpečnostní úřad | ||
309,KVOP,Kancelář veřejného ochránce práv | ||
312,MF,Ministerstvo financí | ||
313,MPSV,Ministerstvo práce a sociálních věcí | ||
314,MV,Ministerstvo vnitra | ||
315,MŽP,Ministerstvo životního prostředí | ||
317,MMR,Ministerstvo pro místní rozvoj | ||
321,GAČR,Grantová agentura ČR | ||
322,MPO,Ministerstvo průmyslu a obchodu | ||
327,MD,Ministerstvo dopravy | ||
328,ČTÚ,Český telekomunikační úřad | ||
329,MZe,Ministerstvo zemědělství | ||
333,MŠMT,"Ministerstvo školství, mládeže a tělovýchovy" | ||
334,MK,Ministerstvo kultury | ||
335,Mzdr,Ministerstvo zdravotnictví | ||
336,Mspr,Ministerstvo spravedlnosti | ||
343,ÚOOÚ,Úřad pro ochranu osobních údajů | ||
344,ÚPV,Úřad průmyslového vlastnictví | ||
345,ČSÚ,Český statistický úřad | ||
346,ČÚZK,Český úřad zeměměřický a katastrální | ||
348,ČBÚ,Český báňský úřad | ||
349,ERÚ,Energetický regulační úřad | ||
353,ÚOHS,Úřad pro ochranu hospodářské soutěže | ||
355,ÚSTR,Ústav pro studium totalitních režimů | ||
358,ÚS,Ústavní soud | ||
361,AVČR,Akademie věd ČR | ||
372,RRTV,Rada pro rozhlasové a televizní vysílání | ||
374,SSHR,Správa státních hmotných rezerv | ||
375,SÚJB,Státní úřad pro jadernou bezpečnost | ||
377,TA ČR,Technologická agentura ČR | ||
381,NKÚ,Nejvyšší kontrolní úřad | ||
376,GIBS,Generální inspekce bezpečnostních sborů | ||
371,ÚDHPSH,Úřad pro dohled nad hospodařením politických stran a politických hnutí | ||
373,ÚPDI,Úřad pro přístup k dopravní infrastruktuře | ||
378,NÚKIB,Národní úřad pro kybernetickou a informační bezpečnost | ||
359,ÚNRR,Úřad Národní rozpočtové rady | ||
362,NSA,Národní sportovní agentura | ||
C E L K E M, Celkem, Celkem |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,21 @@ | ||
rok,url,note,link,filename | ||
2004,https://www.psp.cz/sqw/text/tiskt.sqw?o=4&ct=966&ct1=0,není excel,2004,NA | ||
2005,https://www.psp.cz/sqw/text/tiskt.sqw?o=5&ct=1&ct1=0,není excel,2005,NA | ||
2006,https://www.psp.cz/sqw/text/tiskt.sqw?o=5&ct=198&ct1=0,není excel,2006,NA | ||
2007,https://www.psp.cz/sqw/text/tiskt.sqw?o=5&ct=500&ct1=0,jiná struktura,2007,NA | ||
2008,https://www.psp.cz/sqw/text/tiskt.sqw?o=5&ct=816&ct1=0,není excel,2008,NA | ||
2009,https://www.psp.cz/sqw/text/tiskt.sqw?o=6&ct=1&ct1=0,není excel,2009,NA | ||
2010,https://www.psp.cz/sqw/text/tiskt.sqw?o=6&ct=332&ct1=0,jiná struktura,2010,t0332000017.xls; t0332000018.xls | ||
2011,https://www.psp.cz/sqw/text/tiskt.sqw?o=6&ct=672&ct1=0,NA,2011,NA | ||
2012,https://www.psp.cz/sqw/text/tiskt.sqw?o=6&ct=1010&ct1=0,NA,2012,NA | ||
2013,https://www.psp.cz/sqw/text/tiskt.sqw?o=7&ct=189&ct1=0,NA,2013,NA | ||
2014,https://www.psp.cz/sqw/text/tiskt.sqw?o=7&ct=475&ct1=0,NA,2014,NA | ||
2015,https://www.psp.cz/sqw/text/tiskt.sqw?o=7&ct=800&ct1=0,NA,2015,NA | ||
2016,https://www.psp.cz/sqw/text/tiskt.sqw?o=7&ct=1106&ct1=0,NA,2016,NA | ||
2016,https://www.psp.cz/sqw/text/tiskt.sqw?o=8&ct=1&ct1=0,NA,2016,NA | ||
2017,https://www.psp.cz/sqw/text/tiskt.sqw?o=8&ct=164&ct1=0,NA,2017,NA | ||
2018,https://www.psp.cz/sqw/text/tiskt.sqw?o=8&ct=472&ct1=0,NA,2018,NA | ||
2019,https://www.psp.cz/sqw/text/tiskt.sqw?o=8&ct=843&ct1=0,NA,2019,NA | ||
2020,https://www.psp.cz/sqw/text/tiskt.sqw?o=8&ct=1210&ct1=0,NA,2020,NA | ||
2021,https://www.psp.cz/sqw/text/tiskt.sqw?o=9&ct=212&ct1=0,NA,2021,NA | ||
2022,https://www.psp.cz/sqw/text/tiskt.sqw?o=9&ct=440&ct1=0,NA,2022,t0440f0.xls |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,59 @@ | ||
library(rvest) | ||
library(dplyr) | ||
library(readxl) | ||
library(stringr) | ||
library(readr) | ||
|
||
szu_lst <- readr::read_csv("data-input/szu_tisky-links.csv") | ||
|
||
get_szu_excel <- function(url, filename, year) { | ||
tisk_html <- read_html(url) | ||
|
||
filename_split <- str_split(filename, ";", simplify = FALSE) | ||
filename_clean <- str_trim(filename_split[[1]]) | ||
|
||
yr <- year | ||
|
||
files_tbl <- tibble( | ||
name = tisk_html |> html_elements("td > span.file.xls > a") |> html_text2(), | ||
url = tisk_html |> html_elements("td > span.file.xls > a") |> html_attr("href"), | ||
year = yr | ||
) | ||
|
||
print(filename_clean) | ||
|
||
if(!is.na(filename)) rslt <- files_tbl |> filter(name %in% filename_clean) else rslt <- files_tbl | ||
|
||
return(rslt) | ||
} | ||
|
||
# get_szu_excel(szu_lst$url[10], NA) | ||
# get_szu_excel(szu_lst$url[10], szu_lst$filename[10]) | ||
# get_szu_excel(szu_lst$url[7], szu_lst$filename[7]) | ||
# get_szu_excel(szu_lst$url[20], szu_lst$filename[20]) | ||
|
||
szu_lst_sub <- szu_lst |> select(url, filename, year = rok) |> filter(year > 2009) | ||
|
||
szu_xls_urls <- pmap_dfr(szu_lst_sub, .f = get_szu_excel) |> | ||
mutate(order = row_number(), count = n(), .by = year) | ||
|
||
download_szu_excel <- function(name, url, year, dir) { | ||
outfile <- file.path(dir, paste0(year, "_", name)) | ||
url <- paste0("https://psp.cz", url) | ||
curl::curl_download(url, outfile) | ||
rslt <- tibble(url = url, file = outfile, rok = year) | ||
} | ||
|
||
szu_excel <- pmap_dfr(szu_xls_urls[,1:3], download_szu_excel, "data-input/szu-psp/") | ||
|
||
szu_excel_sheets <- szu_excel |> | ||
mutate(sheets = map(file, readxl::excel_sheets)) |> | ||
unnest(cols = sheets) | ||
|
||
readr::write_rds(szu_excel_sheets, "data-interim/szu-excel-metadata.rds") | ||
|
||
szu_excel_sheets |> | ||
filter(str_detect(sheets, "([Tt]ab1[01])|(Str\\.)")) |> | ||
count(rok, file) | ||
|
||
|
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,91 @@ | ||
library(dplyr) | ||
library(readr) | ||
library(readxl) | ||
library(ggplot2) | ||
|
||
options(scipen = 100) | ||
|
||
szu_excel_sheets <- readr::read_rds("data-interim/szu-excel-metadata.rds") | ||
|
||
read_szu_excel <- function(path, sheet, year) { | ||
is_pocet <- | ||
read_excel(path, sheet, skip = 11) |> | ||
janitor::remove_empty(which = "rows") |> | ||
mutate(rok = year, list = sheet, file = basename(path)) | ||
} | ||
|
||
szu_uo <- szu_excel_sheets |> | ||
mutate(file_hash = tools::md5sum(file)) |> | ||
filter(str_detect(sheets, "ÚO"), str_detect(sheets, "Tab1[10]|Str")) |> | ||
distinct(file_hash, sheets, rok, .keep_all = TRUE) |> | ||
select(path = file, sheet = sheets, year = rok) |> | ||
pmap_dfr(read_szu_excel) |> | ||
janitor::clean_names() |> | ||
group_by(rok, list, file) |> | ||
mutate(promenna = if_else((max(x1_2, na.rm = TRUE) > 1e5), "objemy", "pocty")) |> | ||
ungroup() |> | ||
rename(kap_kod = x1, | ||
kap_zkr = x2, | ||
kap_name = x3, | ||
skut_min = x1_2, | ||
rozp_schv = x2_2, | ||
rozp_pzm = x3_2, | ||
skut = x4, | ||
plneni_proc = x5, | ||
plneni_rozdil = x6, | ||
skut_skut = x7 | ||
) |> | ||
drop_na(kap_kod) |> | ||
mutate(skut_min = if_else(rok < 2013 & promenna == "objemy", skut_min * 1000, skut_min), | ||
rozp_schv = if_else(rok < 2013 & promenna == "objemy", rozp_schv * 1000, rozp_schv), | ||
rozp_pzm = if_else(rok < 2013 & promenna == "objemy", rozp_pzm * 1000, rozp_pzm), | ||
skut = if_else(rok < 2013 & promenna == "objemy", skut * 1000, skut), | ||
kap_name = if_else(is.na(kap_name), kap_zkr, kap_name)) |> | ||
rows_update(read_csv("data-input/kapitoly.csv"), by = "kap_kod") | ||
|
||
szu_uo |> | ||
count(rok, kap_kod, promenna) |> | ||
count(n) | ||
|
||
szu_uo |> | ||
count(rok, wt = is.na(kap_name)) | ||
|
||
szu_uo |> | ||
group_by(rok, kap_kod, promenna) |> | ||
filter(n() > 1) |> | ||
arrange(kap_kod) | ||
|
||
head(szu_uo) | ||
|
||
szu_uo_2009 <- szu_uo |> | ||
filter(rok == 2010) |> | ||
mutate(rok = 2009, | ||
skut = skut_min, | ||
skut_min = NA, | ||
plneni_proc = NA, | ||
plneni_rozdil = NA, | ||
rozp_schv = NA, rozp_pzm = NA, skut_skut = NA) | ||
|
||
szu_uo_wide <- szu_uo |> | ||
bind_rows(szu_uo_2009) |> | ||
pivot_wider(id_cols = c(kap_kod, kap_name, rok, promenna, kap_zkr), | ||
values_from = c(matches("^skut|^rozp|^plneni")), names_from = promenna, names_glue = "{promenna}_{.value}") |> | ||
left_join(df_infl) |> | ||
left_join(wages_later) | ||
|
||
|
||
szu_uo_wide |> | ||
mutate(plat = objemy_skut/pocty_skut/12) |> | ||
filter(plat != 0, kap_zkr != "NSA") |> | ||
ggplot(aes(rok, plat)) + | ||
geom_line() + | ||
geom_line(aes(y = phasal_all), color = "grey") + | ||
facet_wrap(~kap_zkr) | ||
|
||
szu_uo_wide |> | ||
filter(is.na(kap_name)) |> View() | ||
count(kap_kod) | ||
|
||
szu_uo |> | ||
distinct(kap_kod, kap_zkr, kap_name) |> | ||
format_csv() |> cat() |