`createControl`

creates a Cyclops control object for use with `fitCyclopsModel`

.

createControl( maxIterations = 1000, tolerance = 1e-06, convergenceType = "gradient", cvType = "auto", fold = 10, lowerLimit = 0.01, upperLimit = 20, gridSteps = 10, cvRepetitions = 1, minCVData = 100, noiseLevel = "silent", threads = 1, seed = NULL, resetCoefficients = FALSE, startingVariance = -1, useKKTSwindle = FALSE, tuneSwindle = 10, selectorType = "auto", initialBound = 2, maxBoundCount = 5, algorithm = "ccd" )

maxIterations | Integer: maximum iterations of Cyclops to attempt before returning a failed-to-converge error |
---|---|

tolerance | Numeric: maximum relative change in convergence criterion from successive iterations to achieve convergence |

convergenceType | String: name of convergence criterion to employ (described in more detail below) |

cvType | String: name of cross validation search.
Option |

fold | Numeric: Number of random folds to employ in cross validation |

lowerLimit | Numeric: Lower prior variance limit for grid-search |

upperLimit | Numeric: Upper prior variance limit for grid-search |

gridSteps | Numeric: Number of steps in grid-search |

cvRepetitions | Numeric: Number of repetitions of X-fold cross validation |

minCVData | Numeric: Minimum number of data for cross validation |

noiseLevel | String: level of Cyclops screen output ( |

threads | Numeric: Specify number of CPU threads to employ in cross-validation; default = 1 (auto = -1) |

seed | Numeric: Specify random number generator seed. A null value sets seed via |

resetCoefficients | Logical: Reset all coefficients to 0 between model fits under cross-validation |

startingVariance | Numeric: Starting variance for auto-search cross-validation; default = -1 (use estimate based on data) |

useKKTSwindle | Logical: Use the Karush-Kuhn-Tucker conditions to limit search |

tuneSwindle | Numeric: Size multiplier for active set |

selectorType | String: name of exchangeable sampling unit.
Option |

initialBound | Numeric: Starting trust-region size |

maxBoundCount | Numeric: Maximum number of tries to decrease initial trust-region size |

algorithm | String: name of fitting algorithm to employ; default is `ccd` Todo: Describe convegence types |

A Cyclops control object of class inheriting from `"cyclopsControl"`

for use with `fitCyclopsModel`

.

#Generate some simulated data: sim <- simulateCyclopsData(nstrata = 1, nrows = 1000, ncovars = 2, eCovarsPerRow = 0.5, model = "poisson")#> Sparseness = 75.2 %cyclopsData <- convertToCyclopsData(sim$outcomes, sim$covariates, modelType = "pr", addIntercept = TRUE)#> Sorting covariates by covariateId and rowId#Define the prior and control objects to use cross-validation for finding the #optimal hyperparameter: prior <- createPrior("laplace", exclude = 0, useCrossValidation = TRUE) control <- createControl(cvType = "auto", noiseLevel = "quiet") #Fit the model fit <- fitCyclopsModel(cyclopsData,prior = prior, control = control)#> Using cross-validation selector type byRow #> Performing 10-fold cross-validation [seed = 1604638976] with data partitions of sizes 100 100 100 100 100 100 100 100 100 100 #> Using 1 thread(s) #> Starting var = 0.248 (default) #> Running at Laplace(2.83981) None Grid-point #1 at 0.248 Fold #1 Rep #1 pred log like = 380.425 #> Running at Laplace(2.83981) None Grid-point #1 at 0.248 Fold #2 Rep #1 pred log like = 443.495 #> Running at Laplace(2.83981) None Grid-point #1 at 0.248 Fold #3 Rep #1 pred log like = 532.939 #> Running at Laplace(2.83981) None Grid-point #1 at 0.248 Fold #4 Rep #1 pred log like = 539.865 #> Running at Laplace(2.83981) None Grid-point #1 at 0.248 Fold #5 Rep #1 pred log like = 614.918 #> Running at Laplace(2.83981) None Grid-point #1 at 0.248 Fold #6 Rep #1 pred log like = 445.159 #> Running at Laplace(2.83981) None Grid-point #1 at 0.248 Fold #7 Rep #1 pred log like = 464.306 #> Running at Laplace(2.83981) None Grid-point #1 at 0.248 Fold #8 Rep #1 pred log like = 432.631 #> Running at Laplace(2.83981) None Grid-point #1 at 0.248 Fold #9 Rep #1 pred log like = 503.361 #> Running at Laplace(2.83981) None Grid-point #1 at 0.248 Fold #10 Rep #1 pred log like = 366.59 #> AvgPred = 472.369 with stdev = 72.3644 #> Completed at 0.248 #> Next point at 2.48 with value 0 and continue = 1 #> search[ 0.248 ] = 472.369(72.3644) #> #> Running at Laplace(0.898027) None Grid-point #2 at 2.48 Fold #1 Rep #1 pred log like = 380.438 #> Running at Laplace(0.898027) None Grid-point #2 at 2.48 Fold #2 Rep #1 pred log like = 443.499 #> Running at Laplace(0.898027) None Grid-point #2 at 2.48 Fold #3 Rep #1 pred log like = 532.897 #> Running at Laplace(0.898027) None Grid-point #2 at 2.48 Fold #4 Rep #1 pred log like = 539.811 #> Running at Laplace(0.898027) None Grid-point #2 at 2.48 Fold #5 Rep #1 pred log like = 614.929 #> Running at Laplace(0.898027) None Grid-point #2 at 2.48 Fold #6 Rep #1 pred log like = 445.128 #> Running at Laplace(0.898027) None Grid-point #2 at 2.48 Fold #7 Rep #1 pred log like = 464.308 #> Running at Laplace(0.898027) None Grid-point #2 at 2.48 Fold #8 Rep #1 pred log like = 432.602 #> Running at Laplace(0.898027) None Grid-point #2 at 2.48 Fold #9 Rep #1 pred log like = 503.355 #> Running at Laplace(0.898027) None Grid-point #2 at 2.48 Fold #10 Rep #1 pred log like = 366.587 #> AvgPred = 472.356 with stdev = 72.3594 #> Completed at 2.48 #> Next point at 0.0248 with value 0 and continue = 1 #> search[ 0.248 ] = 472.369(72.3644) #> search[ 2.48 ] = 472.356(72.3594) #> #> Running at Laplace(8.98027) None Grid-point #3 at 0.0248 Fold #1 Rep #1 pred log like = 380.307 #> Running at Laplace(8.98027) None Grid-point #3 at 0.0248 Fold #2 Rep #1 pred log like = 443.48 #> Running at Laplace(8.98027) None Grid-point #3 at 0.0248 Fold #3 Rep #1 pred log like = 533.011 #> Running at Laplace(8.98027) None Grid-point #3 at 0.0248 Fold #4 Rep #1 pred log like = 540.025 #> Running at Laplace(8.98027) None Grid-point #3 at 0.0248 Fold #5 Rep #1 pred log like = 614.806 #> Running at Laplace(8.98027) None Grid-point #3 at 0.0248 Fold #6 Rep #1 pred log like = 445.251 #> Running at Laplace(8.98027) None Grid-point #3 at 0.0248 Fold #7 Rep #1 pred log like = 464.311 #> Running at Laplace(8.98027) None Grid-point #3 at 0.0248 Fold #8 Rep #1 pred log like = 432.745 #> Running at Laplace(8.98027) None Grid-point #3 at 0.0248 Fold #9 Rep #1 pred log like = 503.375 #> Running at Laplace(8.98027) None Grid-point #3 at 0.0248 Fold #10 Rep #1 pred log like = 366.597 #> AvgPred = 472.391 with stdev = 72.3688 #> Completed at 0.0248 #> Next point at 0.00248 with value 0 and continue = 1 #> search[ 0.0248 ] = 472.391(72.3688) #> search[ 0.248 ] = 472.369(72.3644) #> search[ 2.48 ] = 472.356(72.3594) #> #> Running at Laplace(28.3981) None Grid-point #4 at 0.00248 Fold #1 Rep #1 pred log like = 379.821 #> Running at Laplace(28.3981) None Grid-point #4 at 0.00248 Fold #2 Rep #1 pred log like = 443.423 #> Running at Laplace(28.3981) None Grid-point #4 at 0.00248 Fold #3 Rep #1 pred log like = 533.208 #> Running at Laplace(28.3981) None Grid-point #4 at 0.00248 Fold #4 Rep #1 pred log like = 540.45 #> Running at Laplace(28.3981) None Grid-point #4 at 0.00248 Fold #5 Rep #1 pred log like = 614.424 #> Running at Laplace(28.3981) None Grid-point #4 at 0.00248 Fold #6 Rep #1 pred log like = 445.486 #> Running at Laplace(28.3981) None Grid-point #4 at 0.00248 Fold #7 Rep #1 pred log like = 464.303 #> Running at Laplace(28.3981) None Grid-point #4 at 0.00248 Fold #8 Rep #1 pred log like = 433.094 #> Running at Laplace(28.3981) None Grid-point #4 at 0.00248 Fold #9 Rep #1 pred log like = 503.29 #> Running at Laplace(28.3981) None Grid-point #4 at 0.00248 Fold #10 Rep #1 pred log like = 366.502 #> AvgPred = 472.4 with stdev = 72.397 #> Completed at 0.00248 #> Next point at 0.000248 with value 0 and continue = 1 #> search[ 0.00248 ] = 472.4(72.397) #> search[ 0.0248 ] = 472.391(72.3688) #> search[ 0.248 ] = 472.369(72.3644) #> search[ 2.48 ] = 472.356(72.3594) #> #> Running at Laplace(89.8027) None Grid-point #5 at 0.000248 Fold #1 Rep #1 pred log like = 379.632 #> Running at Laplace(89.8027) None Grid-point #5 at 0.000248 Fold #2 Rep #1 pred log like = 443.287 #> Running at Laplace(89.8027) None Grid-point #5 at 0.000248 Fold #3 Rep #1 pred log like = 533.459 #> Running at Laplace(89.8027) None Grid-point #5 at 0.000248 Fold #4 Rep #1 pred log like = 541.143 #> Running at Laplace(89.8027) None Grid-point #5 at 0.000248 Fold #5 Rep #1 pred log like = 614.167 #> Running at Laplace(89.8027) None Grid-point #5 at 0.000248 Fold #6 Rep #1 pred log like = 445.322 #> Running at Laplace(89.8027) None Grid-point #5 at 0.000248 Fold #7 Rep #1 pred log like = 464.2 #> Running at Laplace(89.8027) None Grid-point #5 at 0.000248 Fold #8 Rep #1 pred log like = 433.781 #> Running at Laplace(89.8027) None Grid-point #5 at 0.000248 Fold #9 Rep #1 pred log like = 503.079 #> Running at Laplace(89.8027) None Grid-point #5 at 0.000248 Fold #10 Rep #1 pred log like = 366.266 #> AvgPred = 472.434 with stdev = 72.4588 #> Completed at 0.000248 #> Next point at 2.48e-05 with value 0 and continue = 1 #> search[ 0.000248 ] = 472.434(72.4588) #> search[ 0.00248 ] = 472.4(72.397) #> search[ 0.0248 ] = 472.391(72.3688) #> search[ 0.248 ] = 472.369(72.3644) #> search[ 2.48 ] = 472.356(72.3594) #> #> Running at Laplace(283.981) None Grid-point #6 at 2.48e-05 Fold #1 Rep #1 pred log like = 379.632 #> Running at Laplace(283.981) None Grid-point #6 at 2.48e-05 Fold #2 Rep #1 pred log like = 443.287 #> Running at Laplace(283.981) None Grid-point #6 at 2.48e-05 Fold #3 Rep #1 pred log like = 533.459 #> Running at Laplace(283.981) None Grid-point #6 at 2.48e-05 Fold #4 Rep #1 pred log like = 541.143 #> Running at Laplace(283.981) None Grid-point #6 at 2.48e-05 Fold #5 Rep #1 pred log like = 614.167 #> Running at Laplace(283.981) None Grid-point #6 at 2.48e-05 Fold #6 Rep #1 pred log like = 445.322 #> Running at Laplace(283.981) None Grid-point #6 at 2.48e-05 Fold #7 Rep #1 pred log like = 464.2 #> Running at Laplace(283.981) None Grid-point #6 at 2.48e-05 Fold #8 Rep #1 pred log like = 433.781 #> Running at Laplace(283.981) None Grid-point #6 at 2.48e-05 Fold #9 Rep #1 pred log like = 503.079 #> Running at Laplace(283.981) None Grid-point #6 at 2.48e-05 Fold #10 Rep #1 pred log like = 366.266 #> AvgPred = 472.434 with stdev = 72.4588 #> Completed at 2.48e-05 #> Next point at 1.17698e-26 with value 472.6 and continue = 0 #> search[ 2.48e-05 ] = 472.434(72.4588) #> search[ 0.000248 ] = 472.434(72.4588) #> search[ 0.00248 ] = 472.4(72.397) #> search[ 0.0248 ] = 472.391(72.3688) #> search[ 0.248 ] = 472.369(72.3644) #> search[ 2.48 ] = 472.356(72.3594) #> #> #> Maximum predicted log likelihood (472.6) estimated at: #> 1.17698e-26 (variance) #> 1.30356e+13 (lambda) #> #> Fitting model at optimal hyperparameter #> Using prior: Laplace(1.30356e+13) None#> [1] 1.176981e-26#> 'log Lik.' -2056.259 (df=3)#> (Intercept) 1 2 #> -3.850744 0.000000 0.000000#> Using 1 thread(s)#> covariate 2.5 % 97.5 % evaluations #> [1,] 0 -3.877549 -3.824165 24