Skip to contents

Full Information by Exponential Decay

Source: R/full_information.R
exp_decay.Rd

Exponential smoothing twists probabilities by giving relatively more weight to recent observations at an exponential rate.

Usage

exp_decay(x, lambda)

# S3 method for default
exp_decay(x, lambda)

# S3 method for numeric
exp_decay(x, lambda)

# S3 method for matrix
exp_decay(x, lambda)

# S3 method for ts
exp_decay(x, lambda)

# S3 method for xts
exp_decay(x, lambda)

# S3 method for data.frame
exp_decay(x, lambda)

# S3 method for tbl
exp_decay(x, lambda)

Arguments

x

An univariate or a multivariate distribution.

lambda

A double for the decay parameter.

Value

A numerical vector of class ffp with the new probabilities distribution.

Details

The half-life is linked with the lambda parameter as follows:

  • HL = log(2) / lambda.

For example: log(2) / 0.0166 is approximately 42. So, a parameter lambda of 0.0166 can be associated with a half-life of two-months (21 * 2).

See also

crisp kernel_normal half_life

Examples

library(ggplot2)

# long half_life
long_hl <- exp_decay(EuStockMarkets, 0.001)
long_hl
#> <ffp[1860]>
#> 0.0001844669 0.0001846515 0.0001848363 0.0001850212 0.0001852063 ... 0.001183783
autoplot(long_hl) +
  scale_color_viridis_c()


# short half_life
short_hl <- exp_decay(EuStockMarkets, 0.015)
short_hl
#> <ffp[1860]>
#> 1.154879e-14 1.172333e-14 1.19005e-14 1.208036e-14 1.226293e-14 ... 0.01488806
autoplot(short_hl) +
  scale_color_viridis_c()