Summary Function for Bootstrap Inference for Gaussian Finite Mixture Models
summary.MclustBootstrap.RdSummary of bootstrap distribution for the parameters of a Gaussian mixture model providing either standard errors or percentile bootstrap confidence intervals.
Arguments
- object
An object of class
'MclustBootstrap'as returned byMclustBootstrap.- what
A character string:
"se"for the standard errors;"ci"for the confidence intervals;"ave"for the averages.- conf.level
A value specifying the confidence level of the interval.
- ...
Further arguments passed to or from other methods.
Details
For details about the procedure used to obtain the bootstrap distribution see MclustBootstrap.
Examples
# \donttest{
data(diabetes)
X = diabetes[,-1]
modClust = Mclust(X)
bootClust = MclustBootstrap(modClust)
summary(bootClust, what = "se")
#> ----------------------------------------------------------
#> Resampling standard errors
#> ----------------------------------------------------------
#> Model = VVV
#> Num. of mixture components = 3
#> Replications = 999
#> Type = nonparametric bootstrap
#>
#> Mixing probabilities:
#> 1 2 3
#> 0.05280465 0.05190581 0.03553151
#>
#> Means:
#> 1 2 3
#> glucose 1.059243 3.29787 16.83258
#> insulin 7.643779 28.60073 65.85498
#> sspg 7.876315 30.68940 10.19245
#>
#> Variances:
#> [,,1]
#> glucose insulin sspg
#> glucose 11.26049 51.40973 51.09059
#> insulin 51.40973 510.55539 411.22677
#> sspg 51.09059 411.22677 629.55650
#> [,,2]
#> glucose insulin sspg
#> glucose 63.64902 605.8844 435.5397
#> insulin 605.88438 7193.1861 3138.5999
#> sspg 435.53973 3138.5999 6743.2324
#> [,,3]
#> glucose insulin sspg
#> glucose 994.6153 4162.820 644.5863
#> insulin 4162.8201 19343.897 2521.2694
#> sspg 644.5863 2521.269 480.5518
summary(bootClust, what = "ci")
#> ----------------------------------------------------------
#> Resampling confidence intervals
#> ----------------------------------------------------------
#> Model = VVV
#> Num. of mixture components = 3
#> Replications = 999
#> Type = nonparametric bootstrap
#> Confidence level = 0.95
#>
#> Mixing probabilities:
#> 1 2 3
#> 2.5% 0.4458369 0.1560133 0.1321350
#> 97.5% 0.6510765 0.3583681 0.2688954
#>
#> Means:
#> [,,1]
#> glucose insulin sspg
#> 2.5% 89.06346 345.2369 150.4931
#> 97.5% 93.29893 375.9444 182.1779
#> [,,2]
#> glucose insulin sspg
#> 2.5% 99.0615 451.4019 259.9782
#> 97.5% 112.1140 557.2159 380.1439
#> [,,3]
#> glucose insulin sspg
#> 2.5% 198.1075 969.0117 62.12093
#> 97.5% 262.3512 1223.5174 101.47840
#>
#> Variances:
#> [,,1]
#> glucose insulin sspg
#> 2.5% 37.90691 1261.326 1514.129
#> 97.5% 80.82847 3236.033 4086.728
#> [,,2]
#> glucose insulin sspg
#> 2.5% 90.29277 3487.952 12445.55
#> 97.5% 355.76032 30435.133 38721.76
#> [,,3]
#> glucose insulin sspg
#> 2.5% 3330.902 46411.4 1322.555
#> 97.5% 7137.835 121884.5 3225.706
data(acidity)
modDens = densityMclust(acidity, plot = FALSE)
modDens = MclustBootstrap(modDens)
summary(modDens, what = "se")
#> ----------------------------------------------------------
#> Resampling standard errors
#> ----------------------------------------------------------
#> Model = E
#> Num. of mixture components = 2
#> Replications = 999
#> Type = nonparametric bootstrap
#>
#> Mixing probabilities:
#> 1 2
#> 0.04039717 0.04039717
#>
#> Means:
#> 1 2
#> 0.04579252 0.06896503
#>
#> Variances:
#> 1 2
#> 0.02343412 0.02343412
summary(modDens, what = "ci")
#> ----------------------------------------------------------
#> Resampling confidence intervals
#> ----------------------------------------------------------
#> Model = E
#> Num. of mixture components = 2
#> Replications = 999
#> Type = nonparametric bootstrap
#> Confidence level = 0.95
#>
#> Mixing probabilities:
#> 1 2
#> 2.5% 0.5410029 0.3008525
#> 97.5% 0.6991475 0.4589971
#>
#> Means:
#> 1 2
#> 2.5% 4.282648 6.186636
#> 97.5% 4.458571 6.459078
#>
#> Variances:
#> 1 2
#> 2.5% 0.1404628 0.1404628
#> 97.5% 0.2325106 0.2325106
# }