run_all: Imputes the data in all csv-files of directory and writes new...

In MAGDa-BeuthHS/imputeData: Read timeseries from csv-files and impute missing Data

Description

Reads, preprocesses and imputes the data in all csv-files of a directory and writes new files to a subdirectory. Default-strategies are those recommended in the text. Only works for data-aggregation-interval of 5 minutes. Lines in the input-csv files have to be in this format: 182,2014-01-01 00:01:00,0,0 (which means: sensor-id, timestamp, occupancy, speed). Filenames of input-csv: sensor-x.csv (where x can be any number)

Usage

run_all(source_dir = "data/missing", target_dir = "data/imputed",
  vars = c("b", "g"), choices = get_std_choices(),
  pattern = "sensor-[0-9]*.csv", first_timestamp = "2014-01-01 00:00:00",
  last_timestamp = "2014-12-31 23:59:00", header = F, sep = ",",
  append = F)

Arguments

source_dir	the directory of the input-csv-files
target_dir	the directory where the result-csv-files should be written to
vars	vector of the variablenames to be imputed
choices	list of strategies, including gapsize-limits and function-parameters
pattern	pattern of the input-file-names
first_timestamp	first date of the result-data (as character)
last_timestamp	last date of the result-data (as character)
header	logical, if csv-input-file has header: T
sep	the separator between columns in the csv-input
append	logical, if all results should be written into one csv-file: T

Value

nothing, but writes new csv-files with imputed data to target_dir

MAGDa-BeuthHS/imputeData documentation built on May 8, 2019, 9:28 a.m.