merge with remote

[talweg.git] / pkg / R / sample_dataset.R

diff --git a/pkg/R/sample_dataset.R b/pkg/R/sample_dataset.R
new file mode 100644 (file)
index 0000000..6796469
--- /dev/null
+++ b/pkg/R/sample_dataset.R
@@ -0,0 +1,32 @@
+#' Sample dataset to run the algorithm
+#'
+#' Original data from AirNormand was removed, and replaced with the following:
+#' \itemize{
+#'   \item{daily_exogenous.csv}{a CSV file containing dates in the first column, and then n columns for measurements, and finally n columns for predictions (same day)}
+#'   \item{intraday_measures.csv}{a CSV file with two columns: the first with datetimes, and the second with measurements}
+#' }
+#' -----\cr
+#' Here is the code that generated these files: (not any physical sense)\cr
+#' #dts = timestamps from 1970-01-01 01:00 to 1972-01-01 00:00\cr
+#' dts = seq(0,17519*3600,3600)\cr
+#' intraday = data.frame(\cr
+#'   Time=as.POSIXct(dts,tz="GMT",origin="1970-01-01 01:00"),\cr
+#'   Measure=rgamma(length(dts),10,.7) )\cr
+#' dates = seq(as.Date("1970-01-01"), as.Date("1972-01-01"), "day")\cr
+#' m1 = cos(seq_along(dates))\cr
+#' m2 = log(seq_along(dates)+1)\cr
+#' daily = data.frame(\cr
+#'   Date=dates,\cr
+#'   m1=m1,\cr
+#'   m2=m2,\cr
+#'   m1_pred=m1+rnorm(length(m1),sd=.1),\cr
+#'   m2_pred=m2+rnorm(length(m2),sd=.1) )\cr
+#' write.csv(intraday, file="intraday_measures.csv", row.names=F)\cr
+#' write.csv(daily, file="daily_exogenous.csv", row.names=F)
+#'
+#' @name sample
+#' @docType data
+#' @usage data(talweg::sample)
+#' @format Two dataframes: intraday with 17519 rows and 2 columns (Time,Measure), hourly;
+#'   and daily with 731 rows, 5 columns (Date,m1,m2,m1_pred,m2_pred)
+NULL