junos
/
rapids
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

Add label and dates for external files

pull/95/head
Meng Li 2020-02-18 16:41:10 -05:00
parent edcd3da8f9
commit 76888b412d
2 changed files with 9 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
rules/preprocessing.snakefile
View File

@ -1,7 +1,8 @@
rule download_participants:
params:
group = config["DOWNLOAD_PARTICIPANTS"]["GROUP"],
ignored_device_ids = config["DOWNLOAD_PARTICIPANTS"]["IGNORED_DEVICE_IDS"]
ignored_device_ids = config["DOWNLOAD_PARTICIPANTS"]["IGNORED_DEVICE_IDS"],
timezone = config["TIMEZONE"]
script:
"../src/data/download_participants.R"

9
src/data/download_participants.R
View File

@ -4,21 +4,26 @@ library(RMySQL)
group <- snakemake@params[["group"]]
ignored_device_ids <- snakemake@params[["ignored_device_ids"]]
timezone <- snakemake@params[["timezone"]]
rmysql.settingsfile <- "./.env"
stopDB <- dbConnect(MySQL(), default.file = rmysql.settingsfile, group = group)
query <- paste0("SELECT device_id, brand FROM aware_device order by timestamp asc")
query <- "SELECT device_id, brand, label, timestamp FROM aware_device order by timestamp asc"
participants <- dbGetQuery(stopDB, query)
pids <- c()
end_date <- format(Sys.Date(), "%Y/%m/%d")
for(id in 1:nrow(participants)){
device_id <- participants$device_id[[id]]
brand <- ifelse(participants$brand[[id]] == "iPhone", "ios", "android")
label <- ifelse(participants$label[[id]] == "", "EMPTY_LABEL", participants$label[[id]])
start_date <- format(as.POSIXct(participants$timestamp[[id]] / 1000, origin = "1970-01-01", tz = timezone), "%Y/%m/%d")
if(!(device_id %in% ignored_device_ids)){
pid <- paste0("p", ifelse(id < 10, paste0("0", id), id))
pids <- append(pids, pid)
file_connection <- file(paste0("./data/external/", pid))
writeLines(c(device_id, brand), file_connection)
writeLines(c(device_id, brand, label, paste0(start_date, ",", end_date)), file_connection)
close(file_connection)
}
}