This function performs the nonlinear fit of GGamma CDF of a variable x
fitGGammaDist(
x,
parameter.values,
location.par = FALSE,
sample.size = 20,
npoints = NULL,
maxiter = 1024,
ftol = 1e-12,
ptol = 1e-12,
maxfev = 1e+05,
nlms = FALSE,
verbose = TRUE,
...
)numerical vector
initial parameter values for the nonlinear fit. If the locator parameter is included (mu != 0), this must be given as parameter.values = list(alpha = 'value', scale = 'value', mu = 'value', psi = 'value') or if mu = 0, as: parameter.values =list(alpha = 'value', scale = 'value', psi = 'value'). If not provided, then an initial guess is provided.
whether to consider the fitting to generalized gamma distribution (GGamma) including the location parameter, i.e., a GGamma with four parameters (GGamma4P).
Minimum size of the sample.
number of points used in the fit. If the number of points if greater than 10^6, then the fit is automatically set to npoints = 999999. However, the reported values for R.Cross.val, AIC, and BIC are computed taking into account the whole set of points.
positive integer. Termination occurs when the number of iterations reaches maxiter. Default value: 1024.
non-negative numeric. Termination occurs when both the actual and predicted relative reductions in the sum of squares are at most ftol. Therefore, ftol measures the relative error desired in the sum of squares. Default value: 1e-12
non-negative numeric. Termination occurs when the relative error between two consecutive iterates is at most ptol. Therefore, ptol measures the relative error desired in the approximate solution. Default value: 1e-12.
integer; termination occurs when the number of calls to fn has reached maxfev. Note that nls.lm sets the value of maxfev to 100*(length(par) + 1) if maxfev = integer(), where par is the list or vector of parameters to be optimized.
Logical. Whether to return the nonlinear model object
nls.lm. Default is FALSE.
if TRUE, prints the function log to stdout
arguments passed to or from other methods.
Model table with coefficients and goodness-of-fit results:
Adj.R.Square, deviance, AIC, R.Cross.val, and rho, as well as, the
coefficient covariance matrix. If nlms = TRUE, then a list with
nonlinear model object nls.lm is returned.
The script algorithm tries to fit the three-parameter GGamma CDF ('GGamma3P') or the four-parameter GGamma ('GGamma4P') using a modification of Levenberg-Marquardt algorithm implemented in function 'nls.lm' from 'minpack.lm' package that is used to perform the nonlinear fit. Cross-validations for the nonlinear regressions (R.Cross.val) were performed in each methylome as described in reference (1). In addition, Stein's formula for adjusted R squared (rho) was used as an estimator of the average cross-validation predictive power (1).
If the number of values to fit is >10^6, the fitting to a GGamma CDF would be a time consuming task. To reduce the computational time, the option summarized.data' can be set 'TRUE'. If npoint != NULL, the original variable values are summarized into 'npoint' bins and their midpoints are used as the new predictors. In this case, only the goodness-of-fit indicators AIC and R.Cross.val are estimated based on all the original variable x values.
Stevens JP. Applied Multivariate Statistics for the Social Sciences. Fifth Edit. Routledge Academic; 2009.
set.seed(123)
## Fitting GGamma3P
x <- rggamma(3000, alpha = 1.03, psi = 0.75, scale = 1.1)
fitGGammaDist(x)
#>
#> *** Trying nonlinear fit to a generalized 3P Gamma distribution model 3000 values...
#> *** Performing nonlinear regression model crossvalidation...
#> Estimate Std. Error t value Pr(>|t|)) Adj.R.Square
#> alpha 1.0646011 0.001915538 555.7713 0 0.999999998182352
#> scale 1.1382887 0.002872487 396.2729 0
#> psi 0.7383895 0.001898632 388.9061 0
#> rho R.Cross.val DEV
#> alpha 0.999999998178102 0.999951555970568 4.54108853939111e-07
#> scale
#> psi
#> AIC BIC COV.alpha COV.scale
#> alpha -26558.6966173308 -26534.6711470602 3.669287e-06 5.367186e-06
#> scale 5.367186e-06 8.251182e-06
#> psi -3.602364e-06 -5.426154e-06
#> COV.psi COV.mu N model
#> alpha -3.602364e-06 NA 3000 GGamma3P
#> scale -5.426154e-06 NA 3000
#> psi 3.604802e-06 NA 3000
## Fitting GGamma4P
x <- x + 1
fitGGammaDist(x, location.par = TRUE)
#>
#> *** Trying nonlinear fit to a generalized 4P Gamma distribution model 3000 values...
#> *** Performing nonlinear regression model crossvalidation...
#> Estimate Std. Error t value Pr(>|t|)) Adj.R.Square
#> alpha 1.0322905 0.0022133696 466.3887 0 0.999999998586538
#> scale 1.0778304 0.0037442874 287.8600 0
#> mu 0.9981787 0.0000679242 14695.4801 0
#> psi 0.7795591 0.0026167686 297.9091 0
#> rho R.Cross.val DEV
#> alpha 0.999999998582287 0.999955221091228 3.53012009550101e-07
#> scale
#> mu
#> psi
#> AIC BIC COV.alpha COV.scale
#> alpha -26809.4738999022 -26779.442062064 4.899005e-06 8.156174e-06
#> scale 8.156174e-06 1.401969e-05
#> mu 8.893850e-08 1.675482e-07
#> psi -5.736776e-06 -9.779601e-06
#> COV.psi COV.mu N model
#> alpha 8.893850e-08 -5.736776e-06 3000 GGamma4P
#> scale 1.675482e-07 -9.779601e-06 3000
#> mu 4.613696e-09 -1.174772e-07 3000
#> psi -1.174772e-07 6.847478e-06 3000