insp_outlier: Time series outlier inspection

In eric-krishna/DDMEutils: R para o dia a dia

Description

Time series outlier inspection

Usage

insp_outlier(x, ...)

## S3 method for class 'ts'
insp_outlier(x, window_size = 3, anom_method = c("gesd",
  "iqr"))

## S3 method for class 'data.frame'
insp_outlier(x, margin = 1L, window_size = 3,
  .parallel = FALSE, idcol = NULL, dtcol = NULL,
  out_format = c("wide", "long"), anom_method = c("gesd", "iqr"))

Arguments

x	ts or data.frame
window_size	integer window size. Determines how many observations near an outlier are used to calculate the input value.
anom_method	"gesd" or "iqr", argument passed to anomalize::anomalize().
...	Additional arguments for the data.frame case. See details.

Details

• When margin = 1, set idcol to an integer pointing out the index column that identifies the time series;

• When margin = 2, set dtcol to an integer pointing out the date column;

• .parallel TRUE/FALSE if tasks should be parallelized. If TRUE, uses future and furrr packages.

• out_format "wide" or "long", indicating output format. If margin = 2 and there are many time series, out_format = "long" is easier to handle than wide format.

eric-krishna/DDMEutils documentation built on May 30, 2019, 2:09 p.m.