README.md

In matheuscastro43/finiteMix: Finite Mixture Models

Demonstration of package finiteMix:

This notebook demonstrates many finiteMix features.

In actual moment, more than fifty functions are available in finiteMix. Let's use some of them.

First of all, you need the devtools package to install any package from GitHub, install or load it with:

if(!"devtools" %in% installed.packages())
    install.packages(devtools)
require(devtools)

Now, we just need to install finiteMix :)

install_github("matheuscastro43/finiteMix")

Loading finiteMix package..

require(finiteMix)

Loading required package: finiteMix

You has loaded finiteMix library

By Matheus Oliveira de Castro <[e-mail]>

and Gualberto Agamez Montalvo

Remenber, you can use this package in your text, but do not forget to citate us

citation("finiteMix")

To cite package ‘finiteMix’ in publications use:

  CASTRO M. O.; MONTALVO G. S. A. (2021). finiteMix: Finite Mixture
  Models. R package version 0.5.0.

A BibTeX entry for LaTeX users is

  @Manual{,
    title = {finiteMix: Finite Mixture Models},
    author = {CASTRO M. O.; MONTALVO G. S. A.},
    year = {2021},
    note = {R package version 0.5.0},
  }

ATTENTION: This citation information has been auto-generated from the
package DESCRIPTION file and may need manual editing, see
‘help("citation")’.

Finally, let's see some examples:

Generalized Lindley mixtures:

It is possible get density values with "d" function

dglindley_mix(x = 10, pi = c(0.3, 0.7), alpha = c(10, 17), beta = c(1, 3), gamma = c(2, 4), log = FALSE)

0.0375330345533306

density = function(x) dglindley_mix(x, pi = c(0.3, 0.7), alpha = c(10, 17), beta = c(1, 3), gamma = c(2, 4), log = FALSE)

curve(density, 0, 90, lwd = 3, col = "navy")

To get distribuction (or survival, when lower.tail = F) function values, use "p" function

pglindley_mix(q = c(10, 25, 50, 80), pi = c(0.3, 0.7), alpha = c(10, 17), beta = c(1, 3), gamma = c(2, 4), lower.tail = TRUE,
          log = FALSE)

1. 0.137599094879368
2. 0.301738650388426
3. 0.587909657671024
4. 0.978694472766959

distribution = function(q) pglindley_mix(q, pi = c(0.3, 0.7), alpha = c(10, 17), beta = c(1, 3), gamma = c(2, 4), 
                                         lower.tail = TRUE, log = FALSE)

curve(distribution, 0, 90, lwd = 3, col = "navy")

To get quantilic function values, use "q" function

qglindley_mix(p = c(.25, .50, .75, .95), pi = c(0.3, 0.7), alpha = c(10, 17), beta = c(1, 3), gamma = c(2, 4), 
          lower.tail = TRUE, log = FALSE)

1. 13.7886513837911
2. 45.9643806449501
3. 57.5156423436485
4. 73.4206581987979

quantilic = function(p) qglindley_mix(p, pi = c(0.3, 0.7), alpha = c(10, 17), beta = c(1, 3), gamma = c(2, 4),
                                      lower.tail = TRUE, log = FALSE)

curve(quantilic, 0, 1, lwd = 3, col = "navy")

To generate random values, use "r" function

generating = rglindley_mix(n = 1000, pi = c(0.3, 0.7), alpha = c(10, 17), beta = c(1, 3), gamma = c(2, 4))

To estimate the number of components, use g.search

Only using non-parametric criterions

g.search(generating)

$sum
    g = 1     g = 2     g = 3 
476362.69 108193.85  41777.04

$percentageSum
 g = 1  g = 2  g = 3 
  100% 22.71%  8.77%

$plot

Using also parametric criterions

g.search(generating, lim.em = 5, family = "Generalized Lindley")

[==================================================] 100%



$sum
    g = 1     g = 2     g = 3 
476362.69 108193.85  41777.04

$percentageSum
 g = 1  g = 2  g = 3 
  100% 22.71%  8.77%

$AIC
   g = 1    g = 2    g = 3 
9082.564 8264.287 8268.532

$BIC
   g = 1    g = 2    g = 3 
9097.287 8279.010 8283.256

$plot

To estimate the parameters, use "e" function

estimating = eglindley_mix(generating, g = 2, lim.em = 100)

Limit of EM Iterations (100): 
[====                                              ] 8%

Inside finiteMix package, the same functions presented above can be used in mixture models of distributions:

• Contaminated Normal
• Exponential
• Gama
• Generalized Lindley
• Lindley
• Normal
• Poisson
• Rectangular Beta
• Weibull

matheuscastro43/finiteMix documentation built on March 30, 2022, 12:49 p.m.