Pre-divide Eis for two-tier
[openmx:openmx.git] / src / omxExpectationBA81.cpp
2013-08-12 jpritikinPre-divide Eis for two-tier
2013-08-12 jpritikinReorganized the expected weights for cache friendliness
2013-08-11 jpritikinTranspose outcomeProb for better cache behavior
2013-08-11 jpritikinRemove unnecessary memory initialization
2013-08-11 jpritikinFix bug with irregular number of outcomes
2013-08-10 jpritikinTranspose cache to the optimal orientation
2013-08-10 jpritikinRemove estimation of the latent gradient
2013-08-10 jpritikinMove EAP to the new loop nesting order
2013-08-08 jpritikinOptimize E step (helps two-tier more)
2013-08-05 jpritikinMake summary work for IFA models
2013-08-05 jpritikinRemove EItemParam
2013-08-05 jpritikinUse full range of double
2013-08-04 jpritikinAuto-tune Newton-Raphson
2013-08-01 jpritikinFactor out density multiplication
2013-08-01 jpritikinReorg E step 9/9
2013-08-01 jpritikinReorg E step 8/9
2013-08-01 jpritikinReorg E step 7/9
2013-08-01 jpritikinReorg E step 6/9
2013-08-01 jpritikinReorg E step 5/9
2013-08-01 jpritikinReorg E step 4/9
2013-08-01 jpritikinReorg E step 3/9
2013-08-01 jpritikinReorg E step 2/9
2013-08-01 jpritikinReorg E step 1/9
2013-08-01 jpritikinAdd cache for outcomeProb
2013-08-01 jpritikinMove allocation out of computeRPF
2013-08-01 jpritikinStore outcome probs more compactly
2013-07-29 jpritikinRename
2013-07-29 jpritikinFix diag
2013-07-29 jpritikinRemove artificial lower bound on patternLik
2013-07-29 jpritikinFix incorrect handling of missing data
2013-07-29 jpritikinMake more internal data available from R
2013-07-28 jpritikinPermit non-integral quadrature widths (oops)
2013-07-26 jpritikinFail if an outcome is entirely unendorsed
2013-07-26 jpritikinReport the empirical mean & cov
2013-07-22 jpritikinMight as well cache the number of outcomes per item
2013-07-18 jpritikinA few more minor optimizations
2013-07-18 jpritikinProper dependency tracking for BA81
2013-07-18 jpritikinAdd verbose option to BA81 stuff
2013-07-18 jpritikinMove cai2010 data back to simple arrays from std::vector
2013-07-18 jpritikinUse larger buffers to avoid atomic-add synchronization
2013-07-18 jpritikinOptimize estimation of the latent distribution
2013-07-18 jpritikinRemove a few more exp()
2013-07-18 jpritikinAvoid exp(); more than doubles performance; no detectab...
2013-07-18 jpritikinMinimize calls to omx_absolute_thread_num()
2013-07-16 jpritikinMove all fitfunction args to expectation
2013-07-16 jpritikinPass design as a simple matrix, not MxMatrix
2013-07-16 jpritikinGradients for the latent distribution
2013-07-12 jpritikinPass item models in a list instead of an MxMatrix
2013-07-12 jpritikinImprove EAP performance
2013-07-12 jpritikinReduce rpf evaluations in two-tier models
2013-07-11 jpritikinLook at expectation to determine which fit to calculate
2013-07-11 jpritikinAdjust BA81 expectation context argument
2013-07-09 jpritikinMultigroup two-tier
2013-07-08 jpritikinUse correct density for specific dimensions
2013-07-08 jpritikinReport EAP scores organized in the same way as flexMIRT
2013-07-08 jpritikinUpdate diagnostics
2013-07-07 jpritikinTest multigroup, multidimensional EAP
2013-07-06 jpritikinMove EAP scores to expectation slot scores.out
2013-07-06 jpritikinMultigroup IFA
2013-07-06 jpritikinReplace mxComputeAssign with mxAlgebra(..., fixed=TRUE)
2013-07-06 jpritikinNo need to report EM.LL in a special way
2013-07-06 jpritikinDon't round quadrature points
2013-07-06 jpritikinSync libifa-rpf with RPF 0.10
2013-06-28 jpritikinSplit BA81 into a fit function and expectation
2013-06-28 jpritikinFix compile errors after merge
2013-06-28 jpritikinEliminate Rprintf
2013-06-28 jpritikinRename