Skip to content
GitLab
Select Git revision
  • master default
Find file BlameHistoryPermalink
extract.R 10.33 KiB
1 
# Usefull library
2 
library(tools)
3 
library(dplyr)
4 
library(officer)
5
6
7 
# General information on station
8 
iStatut = c('0'='inconnu',
9 
            '1'='station avec signification hydrologique',
10 
            '2'='station sans signification hydrologique',
11 
            '3'="station d'essai")
12
13 
iFinalite = c('0'='inconnue',
14 
              '1'="hydrométrie générale",
15 
              '2'='alerte de crue',
16 
              '3'="hydrométrie générale et alerte de crue",
17 
              '4'="gestion d'ouvrage",
18 
              '5'='police des eaux',
19 
              '6'="suivi d'étiage",
20 
              '7'='bassin expérimental',
21 
              '8'='drainage')
22
23 
iType = c('0'='inconnu',
24 
          '1'='une échelle',
25 
          '2'='deux échelles, station mère',
26 
          '3'='deux échelles, station fille',
27 
          '4'='débits mesurés',
28 
          '5'='virtuelle')
29
30 
iInfluence = c('0'='inconnue',
31 
               '1'='nulle ou faible',
32 
               '2'='en étiage seulement',
33 
               '3'='forte en toute saison')
34
35 
iDebit = c('0'='reconstitué',
36 
           '1'="réel (prise en compte de l'eau rajoutée ou retirée du bassin selon aménagements)",
37 
           '2'='naturel')
38
39 
iQBE = c('0'='qualité basses eaux inconnue',
40 
         '1'='qualité basses eaux bonne',
41 
         '2'='qualité basses eaux douteuse')
42
43 
iQME = c('0'='qualité moyennes eaux inconnue',
44 
         '1'='qualité moyennes eaux bonne',
45 
         '2'='qualité moyennes eaux douteuse')
46
47 
iQHE = c('0'='qualité hautes eaux inconnue',
48 
         '1'='qualité hautes eaux bonne',
49 
         '2'='qualité hautes eaux douteuse')
50
51
52 
# Get the selection of data from the 'Liste-station_RRSE' file and the BanqueHydro directory
53 
get_selection = function (computer_data_path, listdir, listname,
54 
                          cnames=c('code','station', 'BV_km2', 'axe_principal_concerne', 'longueur_serie', 'commentaires', 'choix'),
55 
                          cisnum=c('BV_km2', 'longueur_serie')) {
56
57 
    # Get the file path to the data
58 
    list_path = file.path(computer_data_path, listdir, listname)
59
60 
    # Extract the data as a data frame
61 
    # df_list = read.table(list_path,
62 
    #                      header=TRUE,
63 
    #                      sep=';',
64 
    #                      dec=',',
65 
    #                      quote='',
66 
    #                      skip=0,
67 
    #                      nrows=3,
68 
    #                      strip.white=TRUE,
69 
    #                      comment.char="",
70 
    #                      colClasses=c("character",
7172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140
# "character", # "numeric", # "character", # "numeric", # "character", # "character")) sample_data = read_docx(list_path) content = docx_summary(sample_data) table_cells <- content %>% filter(content_type == "table cell") table_data <- table_cells %>% filter(!is_header) %>% select(row_id, cell_id, text) # Split data into individual columns splits <- split(table_data, table_data$cell_id) splits <- lapply(splits, function(x) x$text) # Combine columns back together in wide format df_list <- bind_cols(splits) df_list = df_list[-1,] names(df_list) = cnames for (c in cisnum) { df_list$c = as.numeric(sub(",", ".", pull(df_list, c))) } df_selec = df_list[df_list$choix == 'A garder' | df_list$choix == 'Ajout',] df_selec = bind_cols(df_selec, filename=paste(df_selec$code, '_HYDRO_QJM.txt', sep='')) return (df_selec) } # Example # df_selec = get_selection( # "/home/louis/Documents/bouleau/INRAE/CDD_stationnarite/data", # "liste_station", # "Liste-station_RRSE.docx", # cnames=c('code','station', # 'BV_km2', # 'axe_principal_concerne', # 'longueur_serie', # 'commentaires', # 'choix'), # cisnum=c('BV_km2', # 'longueur_serie')) # Extraction of information extract_info = function (computer_data_path, filedir, filename, verbose=TRUE) { # Convert the filename in vector filename = c(filename) # If the filename is 'all' or regroup more than one filename if (all(filename == 'all') | length(filename) > 1) { # If the filename is 'all' if (all(filename == 'all')) { # Create a filelist to store all the filename filelist = c() # Get all the filename in the data directory selected filelist_tmp = list.files(file.path(computer_data_path, filedir)) # For all the filename in the directory selected for (f in filelist_tmp) { # If the filename extention is 'txt' if (file_ext(f) == 'txt') {
141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210
# Store the filename in the filelist filelist = c(filelist, f) } } # If the filename regroup more than one filename } else if (length(filename > 1)) { # The filelist correspond to the filename filelist = filename } # Create a blank data frame df_info = data.frame() # For all the file in the filelist for (f in filelist) { # Concatenate by raw data frames created by this function when filename correspond to only one filename df_info = rbind(df_info, extract_info(computer_data_path, filedir, f)) } # Set the rownames by default (to avoid strange numbering) rownames(df_info) = NULL return (df_info) } # Get the filename from the vector filename = filename[1] # Print information if asked if (verbose) { print(paste("extraction of info for file :", filename)) } # Get the file path to the data file_path = file.path(computer_data_path, filedir, filename) # Extract all the header infotxt = c(readLines(file_path, n=41)) # Create a tibble with all the information needed df_info = tibble(code=trimws(substr(infotxt[11], 38, nchar(infotxt[11]))), nom=trimws(substr(infotxt[12], 39, nchar(infotxt[12]))), territoire=trimws(substr(infotxt[13], 39, nchar(infotxt[13]))), L93X=as.numeric(substr(infotxt[16], 38, 50)), L93Y=as.numeric(substr(infotxt[16], 52, 63)), surface=as.numeric(substr(infotxt[19], 38, 50)), statut=iStatut[trimws(substr(infotxt[26], 38, 50))], finalite=iFinalite[trimws(substr(infotxt[26], 52, 56))], type=iType[trimws(substr(infotxt[26], 58, 58))], influence=iInfluence[trimws(substr(infotxt[26], 60, 60))], debit=iDebit[trimws(substr(infotxt[26], 62, 62))], QBE=iQBE[trimws(substr(infotxt[26], 72, 72))], QME=iQME[trimws(substr(infotxt[26], 74, 74))], QHE=iQHE[trimws(substr(infotxt[26], 76, 76))], file_path=file_path ) return (df_info) } # Example # df_info = extract_info( # "/home/louis/Documents/bouleau/INRAE/CDD_stationnarite/data", # '', # c('H5920011_HYDRO_QJM.txt', 'K4470010_HYDRO_QJM.txt'))
211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280
# Extraction of data extract_data = function (computer_data_path, filedir, filename, verbose=TRUE) { # Convert the filename in vector filename = c(filename) # If the filename is 'all' or regroup more than one filename if (all(filename == 'all') | length(filename) > 1) { # If the filename is 'all' if (all(filename == 'all')) { # Create a filelist to store all the filename filelist = c() # Get all the filename in the data directory selected filelist_tmp = list.files(file.path(computer_data_path, filedir)) # For all the filename in the directory selected for (f in filelist_tmp) { # If the filename extention is 'txt' if (file_ext(f) == 'txt') { # Store the filename in the filelist filelist = c(filelist, f) } } # If the filename regroup more than one filename } else if (length(filename > 1)) { # The filelist correspond to the filename filelist = filename } # Create a blank data frame df_data = data.frame() # For all the file in the filelist for (f in filelist) { # Concatenate by raw data frames created by this function when filename correspond to only one filename df_data = rbind(df_data, extract_data(computer_data_path, filedir, f)) } # Set the rownames by default (to avoid strange numbering) rownames(df_data) = NULL return (df_data) } # Get the filename from the vector filename = filename[1] # Print information if asked if (verbose) { print(paste("extraction of data for file :", filename)) } # Get the file path to the data file_path = file.path(computer_data_path, filedir, filename) # Extract the data as a data frame df_data = read.table(file_path, header=TRUE, na.strings=c(' -99', ' -99.000'), sep=';', skip=41)[,1:2] # Extract all the information for the station
281282283284285286287288289290291292293294295296297298
df_info = extract_info(computer_data_path, filedir, filename, verbose=FALSE) # Get the code of the station code = df_info$code # Create a tibble with the date as Date class and the code of the station df_data = tibble(Date=as.Date(as.character(df_data$Date), format="%Y%m%d"), df_data[-1], code=code) return (df_data) } # Example # df_data = extract_data( # "/home/louis/Documents/bouleau/INRAE/CDD_stationnarite/data", # '', # c('H5920011_HYDRO_QJM.txt', 'K4470010_HYDRO_QJM.txt'))