scale_gran: robust_scaling of data marginal on categories of different...

In Sayani07/gracsr:

Description

robust_scaling of data marginal on categories of different granularities

Usage

scale_gran(.data, response = NULL, method = "none")

Arguments

.data	a tsibble
response	measured variable
method	scaling method like "robust" (Robust sscaling), "nqt" (Normal-quantile transform)
gran1	one granularity e.g. hour_day, day_week, wknd_wday
gran2	one granularity distinct from gran1

Value

scale_gran

Examples

library(gravitas)
library(tidyverse)
sm <- smart_meter10 %>%
  filter(customer_id %in% c("10006704", "10017936", "10006414", "10018250"))
gran1 <- "hour_day"
gran2 <- NULL
response <- "general_supply_kwh"
scale_gran(sm, response, method = "none")
scale_gran(sm, response, method = "robust")
scale_gran(sm, response, method = "nqt")

# todo : can add NQT marginal on each category?

Sayani07/gracsr documentation built on Dec. 18, 2021, 12:59 p.m.