line_protocol: Convert data to InfluxDB line protocol

In imlijunda/rinfluxdb: InfluxDB Driver for R

Description

Convert data to InfluxDB line protocol

Usage

line_protocol(data, measurement, tag = NULL, field = NULL,
  time = NULL, chk_escape = FALSE, chk_na = FALSE, use_int = FALSE,
  fp_prec = NULL, epoch = c("ns", "u", "ms", "s", "m", "h"))

line_protocol_gen(ref, measurement, tag = NULL, field = NULL,
  time = NULL, chk_escape = FALSE, chk_na = FALSE, use_int = FALSE,
  fp_prec = NULL, epoch = c("ns", "u", "ms", "s", "m", "h"))

## Default S3 method:
line_protocol_gen(ref, measurement, tag = NULL,
  field = NULL, time = NULL, chk_escape = FALSE, chk_na = FALSE,
  use_int = FALSE, fp_prec = NULL, epoch = c("ns", "u", "ms", "s",
  "m", "h"))

## S3 method for class 'xts'
line_protocol_gen(ref, measurement, tag = NULL,
  field = NULL, time = NULL, chk_escape = FALSE, chk_na = FALSE,
  use_int = FALSE, fp_prec = NULL, epoch = c("ns", "u", "ms", "s",
  "m", "h"))

Arguments

data	data to convert.
measurement	measurement name, or column name of measurement in ref
tag	column names of tags in ref
field	column names of fields in ref, if NULL, all columns will be used as fields
time	column name of time in ref
chk_escape	whether to check special characters to escape
chk_na	whether to check NA values
use_int	whether to use integer data type
fp_prec	precision of floating point data, if NULL, system default is used
epoch	unit of epoch
ref	a reference data structure

Details

line_protocol() converts a named data structure to a character vector of InfluxDB line protocol. It is a wrapper of line_protocol_gen(), which parses reference data and returns a converter function that accepts same data structure.

Remarks on NA values. if chk_na is TURE, line_protocol_gen() checks for NA values in fields ONLY and uses regex ",(?:(?!,|=).)*?=NA" to match/remove NA fields. However this comes with two caveat: "=NA" inside string field values are replace by random token @_=_N_A@ + a random number and later replaced back to =NA, if you have such values and accidentally with same random token, Secondly, as you may notice, the mentioned regex can't distinguish field keys with special characters such as special\,key\=NA=100 will cause problem.

Value

a function that accepts same data structure as ref, or a character vector of line protocol

Examples

ref <- iris
f <- line_protocol_gen(ref, "iris", tag = "Species")
lp <- f(ref)
lp2 <- line_protocol(ref, "iris", tag = "Species")
stopifnot(all(lp == lp2))

imlijunda/rinfluxdb documentation built on Nov. 4, 2019, 2:09 p.m.