quant_bin_1d: One-Dimensional Empirical Quantile Binning

In kmaurer/binsemble:

Description

used for binning the numeric values by empirical quantile

Usage

quant_bin_1d(xs, nbin, output = "data", jit = 0)

Arguments

xs	Numeric vector of values to be binned
nbin	An integer defining the number of bins to partion the xs into
output	Output Structure: "data" for just the binned data,"definition" for a list of bin centers and boundaries, or "both" for list containing both data and definition
jit	non-negative value to specify a random uniform jitter to the observed values prior to partitioning by quantiles.

Value

output as specified

Examples

quant_bin_1d(ggplot2::diamonds$price,4,output="data")
quant_bin_1d(ggplot2::diamonds$price,4,output="definition")
quant_bin_1d(runif(1000,0,10),nbin=4,output="both")

Speed test
load("~/onePercentSample.Rdata")
timer <- Sys.time()
quant_bin_1d(onePercentSample$total_amount,100,output="data", jit=.00001)
Sys.time()-timer
Note: using .bincode() this take ~2 seconds instead of ~10 seconds with for loop overwrite from original

kmaurer/binsemble documentation built on May 7, 2019, 9:50 p.m.