explode: expland out a two dimensional object

In AndreMikulec/econModel: Meaningful Social and Economic Data from ALFRED and elsewhere

Description

From one(1) or two(2) two dimension objects, apply a function(3) upon the two objects. Return a two dimensional object of with a column name derived from the three. Functions are meant to be many of the functions from R CRAN packages package TTR and PerformanceAnalytics. Note, if any x2, then x1 and x2 are paired/matched column position to column position. Note, this the "multivariate form" of the "single variate form" object TTR function data generator from the web page "Time series cross-validation 5".

Usage

explode(
  x1 = NULL,
  x2 = NULL,
  Fun = NULL,
  Flags = NULL,
  AltName = NULL,
  asIsAltName = NULL,
  Prefix = NULL,
  FixedSep = NULL,
  quote = FALSE,
  envir = parent.frame(2),
  ...
)

Arguments

x1	two dimensional object
x2	Optionally, second two dimensional object.
Fun	function name in the "bare" or in literal quotes("")
Flags	list of possible varying parameters that are expanded to all possible combinations by expand.grid
AltName	string alternate name for "Fun"
asIsAltName	Default is NULL(F). Passed to the column generator function. NULL is interpreted as F. This NULL(F) means do not attempt to simplify the AltName. For example, if the user passed "TTR::SMA" and asIsAltName == T, then "TTR::SMA" would directly tried to become part of the column name (and "::" would cause an error so this would not work.). If the user passed "TTR::SMA" and asIsAltName == NULL(F), then "SMA" would be directly tried to directly become part of the column name.
Prefix	boolan default is FALSE. TRUE would place the column meta before the column name.
FixedSep	string divider of meta items
quote	boolean passed to package DescTools function DoCall
envir	calling environment
...	additional parameters

Value

two dimensional object with new and different columns

References

Zachary Mayer, "Time series cross-validation 5" http://www.r-bloggers.com/time-series-cross-validation-5/ http://moderntoolmaking.blogspot.com/2013/01/time-series-cross-validation-5.html

Examples

## Not run: 

library(quantmod)
ibm <- getSymbols("IBM", from = "1970-01-01", to = "1970-01-13", auto.assign = FALSE)

explode(ibm[,c("IBM.Open")], Fun = "TTR::SMA", Flags = list(n = 2:3))
          IBM.Open.TTR.SMA.n.2  IBM.Open.TTR.SMA.n.3
1970-01-02                   NA                   NA
1970-01-05               18.262                   NA
1970-01-06               18.356               18.312
1970-01-07               18.419               18.379
1970-01-08               18.431               18.425
1970-01-09               18.456               18.446
1970-01-12               18.463               18.454

explode(ibm[,c("IBM.Open","IBM.Close")], Fun = "TTR::SMA", Flags = list(n = 2:3))

           IBM.Open.TTR.SMA.n.2 IBM.Close.TTR.SMA.n.2 IBM.Open.TTR.SMA.n.3
1970-01-02                   NA                    NA                   NA
1970-01-05               18.262                18.325                   NA
1970-01-06               18.356                18.419               18.312
1970-01-07               18.419                18.431               18.379
1970-01-08               18.431                18.456               18.425
1970-01-09               18.456                18.463               18.446
1970-01-12               18.463                18.419               18.454

# column naming checks

explode(ibm[,c("IBM.Open","IBM.Close")], Fun = TTR::SMA, Flags = list(n = 2:3),
        AltName = "CUST")
IBM.Open.CUST.n.2 IBM.Close.CUST.n.2 IBM.Open.CUST.n.3 IBM.Close.CUST.n.3

explode(IBM.Open.TTR.SMA.n.2, Fun = TTR::SMA, Flags = list(n = 3:4))
IBM.Open.TTR.SMA.n.2.TTR.SMA.n.3 IBM.Open.TTR.SMA.n.2.TTR.SMA.n.4

SMA2 <- TTR::SMA
explode(IBM.Open.TTR.SMA.n.2, Fun = SMA2, Flags = list(n = 3:4))
IBM.Open.TTR.SMA.n.2.SMA2.n.3 IBM.Open.TTR.SMA.n.2.SMA2.n.4

# x2 case

# R CRAN Package TTR function runCor
# runCor : function (x, y, n = 10, use = "all.obs", sample = TRUE, cumulative = FALSE)
explode(ibm[,c("IBM.Open","IBM.Close")], ibm[,c("IBM.Low","IBM.High")],
        Fun = "TTR::runCor", Flags = list(n = 4:5, sample = c(TRUE,FALSE)))


## End(Not run)

AndreMikulec/econModel documentation built on June 30, 2021, 9:48 a.m.