Release 2.0 (Nov 15, 2026)
====================================
* Release to CRAN.
Breaking changes below are relative to CRAN version 1.5.0.
* Breaking change: R 4.2.0 or later is required; version 1.5.0 supported R 3.4.0 and later.
* Breaking change: variable names, data column names, and ordinary parameter labels containing periods are rejected. Rename them consistently (for example, x.1 to x_1); OpenMx definition-variable references such as data.x_1 remain supported.
* Breaking change: osmasem() has a new RE.User argument before data. Calls passing data or later arguments by position must name those arguments.
* Breaking change: osmasem2() no longer provides likelihood-based confidence intervals or returns the intervals.type field. Remove intervals.type; a likelihood-based request now raises an error and z is deprecated. Arguments after mean.analysis in older positional calls must be named.
* Breaking change: is.pd() now returns FALSE for singular or asymmetric matrices previously accepted as positive definite.
* Breaking change: list2matrix(), tssem1(), and Cor2DataFrame() require study matrices either to have consistent, unique row and column variable names or to be entirely unnamed. osmasem() and osmasem2() reject partially matching RAM and data variable names; rename them consistently.
* Breaking change: rCorSam(), asyCov(), and asyCovOld() require matching input lengths unless one input has length one; asyCov sample sizes must be positive and finite. smdMTS(homogeneity="none", all.comparisons=TRUE) rejects an incomplete first group instead of using another group as the standardizer.
* Breaking change: osmasem() rejects covariance-formatted data; use osmasem2(cor.analysis=FALSE) for covariance models.
* Breaking change: summary() for sem() fits rejects robust=TRUE with likelihood-based intervals; use intervals.type="z" for robust standard errors or robust=FALSE for likelihood-based intervals.
* Breaking change: calEffSizes() now suppresses lavaan warnings by default; use warn=TRUE to show them as in version 1.5.0.
* Breaking change: lavaan2RAM() is restricted to documented options and syntax. Composite `<~` and pass-through options such as orthogonal, group.equal, and effect.coding are no longer supported; use explicit group labels such as c(b, b)*x for equality. Some previously accepted options could silently produce the wrong model.
* Breaking change: meta() and metaFIML() reject missing sampling covariances between observed effects; sem() rejects unnamed or unknown bounds.
* Breaking change: asyCov(cor.analysis=FALSE) returns NA for sampling covariances that depend on missing input covariances; previous sampling weights and fitted results could be incorrect.
* Breaking change: asyCov() rejects non-positive-definite working matrices; asyCovOld() more strictly checks positive definiteness and rejects missing covariances among retained variables. The legacy tssem1() path no longer replaces missing observed covariances with zero.
* Breaking change: tssem1() rejects reported statistics without available sampling variances rather than silently omitting them; partial working-matrix validation now systematically checks observed submatrices.
* Breaking change: asyCov() rejects asymmetric missingness and non-finite observed entries.
* Breaking change: as.mxMatrix() preserves supplied dimension names, and generated multi-column moderator names use underscores; clustered tssem1FEM() now honors run=FALSE.
* Extend sem() to support combined multiple-group and two-level models, cluster-level covariates, and between-level definition variables.
* Add cluster-robust standard errors and clusterRobustSE(), and t-based tests and intervals for sem() fits.
* Add SEC() and CCA() for overlapping meta-analyses, and the Chan25 and Turk23 datasets.
* Improve handling of constraints in sem(), lavaan2RAM(), and osmasem2(), including tighter feasibility tolerance and corrected standard errors. In osmasem2() correlation models, constrain observed total variances even when residual variances are fixed. In osmasem(), retain RAM bounds where A.lbound or A.ubound entries are NA. These changes can affect results.
* Warn when correlation models may ignore fixed dependent residual variances or factor scales may be unanchored; correct lavaan2RAM() default mean starts and labels.
* Correct lavaan2RAM() indicator intercepts when a latent mean is explicitly fixed; previous estimates and fit could be wrong. Preserve small variances in two-level summary input, which could otherwise change estimates and likelihood.
* Match named group and cluster summary inputs by name in sem(); previously mismatched orders could change estimates. Correct two-level sample and cluster counts, warn about unidentified latent means, and fix combined-model submodel name collisions and auxiliary-model summaries.
* Fix covariance-only likelihood-based intervals in sem() and improve the accuracy of SEM diagrams, including variable names and fixed/free paths.
* Correct lavaan2RAM() two-level default means: within-only observed variables are no longer fixed at zero, and an explicit intercept no longer fixes other default means at zero. Correct sem() level-1 paths to use the within component of predictors that also appear at level 2. These fixes can change estimates and fit.
* Pair lavaan2RAM() group blocks with data in block order, including labeled groups; previously, models could be assigned to the wrong groups. Apply parameter-to-constant inequalities as bounds instead of dropping them.
* Correct osmasem() and osmasem2() to match RAM observed variables to data by name. Previously, different variable orders could associate estimates with the wrong paths; for example, the Roorda pos and neg paths could be swapped. When no names match, positional matching is retained with a warning.
* Correct variable ordering for direct osmasem(Mmatrix=...) input and path names in wls() and osmasem() plots. Direct-input estimates may change.
* Match named RAM observed variables to the pooled matrix in wls() and tssem2(); different orders previously could produce incorrect estimates and fit. Correct subset.variables ordering of the asymptotic covariance matrix; this can also change previous results.
* Correct uniR2mx() variable alignment; different RAM and pooled-matrix orders could previously misassign estimates, especially with unnamed F matrices.
* Align named correlation matrices by variable name in tssem1() and Cor2DataFrame(); this can affect existing results.
* Improve convergence and diagnostic warnings, including clear errors for fixed-effects TSSEM studies without observed correlations or variances; retry sem() and osmasem() fits with mxTryHard() when needed. Reject duplicate group or cluster columns in combined sem() summary input.
* Correct pattern.na(type="osmasem", show.na=FALSE) off-diagonal counts to count studies with both correlations observed, rather than at least one. Diagonal counts and type="tssem" are unaffected; the correction changes the diagnostic table, not osmasem() model fitting.
* Correct smdMTS() and smdMES() bias adjustments for the standardizer actually used; adjusted estimates and sampling covariances can change. Zero means are retained.
* Correct osmasem() covariance-moderator calculations and fixed-effect extraction; estimates can change.
* Align named sampling covariances and separately supplied filter matrices in wls(), and named study matrices in asyCov(); previous estimates can change.
* Preserve bounds when substituting equality constraints and avoid automatic parameter-label collisions; these changes can affect results. Generated correlation names are made unique when underscores would otherwise cause collisions.
* Correct multigroup SEM name collisions, WLS pair matching, clustered TSSEM arguments, and CFI at the zero-noncentrality boundary.
* Correct the Becker09 and Roorda11 random-effects examples to retain reported correlations whose unit diagonals were missing; use implicit diagonal constraints in the Roorda11 stage-two example.
* Correct fit statistics for zero-degree-of-freedom models.
* Correct fixed-effects TSSEM baseline likelihoods and use total sample size N for RMSEA estimates and confidence intervals in fixed-effects TSSEM and WLS; chi-square and related fit statistics can change. Retain variance-only studies in covariance analyses.
* Build the vignettes from source and update their narratives to use fitted results.
* Restore function and method usage information in the help pages.
* Improve plotting and fix other issues in SEM and meta-analysis functions.

Release 1.6.0 (Aug 22, 2026)
====================================
* Extend lavaan2RAM() and sem() to fit multiple-group and two-level models specified with lavaan syntax.

Release 1.5.6 (Jul 18, 2026)
====================================
* Rewrite homoStat to improve speed and RAM use.

Release 1.5.5 (May 02, 2026)
====================================
* Use roxygen2 for documentation.

Release 1.5.4 (Feb 10, 2026)
====================================
* Extend the RE.type argument in osmasem() to include "CSymm" and "HCSymm" options.

Release 1.5.3 (Oct 18, 2025)
====================================
* Update sem() and osmasem2().
* Add error and warning messages in mxsem and osmasem2 objects.

Release 1.5.0 (Sep 26, 2024)
====================================
* Add dataset Chan17.
* Add lbound and ubound in sem() and as.mxAlgebra().
* Add a field of citations in the Jaramillo05 dataset.
* Rename create.mxModel() to sem().
* Add osmasem2() and friends.
* Add mean data in issp05.
* Support means in Cor2DataFrame().
* Fix the means of the latent variables at 0 in lavaan2RAM().
* Add "<~" operator in lavaan2RAM().
* Fix startvalues not assigned in create.mxModel().

Release 1.4.0 (May 15, 2024)
====================================
* Release to CRAN.
* Revise create.mxModel().
* Replace dontrun with donttest.

Release 1.3.1 (Aug 8, 2023)
====================================
* Release to CRAN.
* Lower the tolerance in testing asyCov(), which returns an error in latest R.

Release 1.3.0 (Jan 7, 2023)
====================================
* Fix a bug reported by Noel Card that meta() reported incorrect number of statistics and dfs.
* Rewrite asyCov() and rename the old one to asyCovOld().
* Add rCor3L() and its summary method.
* Add subset.variables argument in tssem2() and wls().
* Rename meta3() to meta3L() and meta3X() and meta3LFIML().
* Fix a critical bug in create.Tau2() that uses variances as SDs.
* Add a new argument 'level' in create.Tau2().
* Add a mxModel.Args in wls().
* Add A.lbound and A.ubound arguments in create.vechsR() and osmasem().
* Add data and group arguments in calEffSizes().
* Modify Nohe15A1 and Nohe15A2.
* Add a RAM argument in impliedR() and rimpliedR() and an SmatrixSD argument in rimpliedR(). 
* Add a starting value argument in as.mxAlgebra().
* Teach lavaan2RAM() to recognize functions of parameters and constraints in lavaan and convert them to mxAlgebra and mxConstraint in OpenMx.
* Experimental: teach create.mxModel() to recognize mxAlgrebra and mxConstraint converted from lavaan2RAM().

Release 1.2.5.1 (May 17, 2021)
====================================
* Release to CRAN.
* Fix a bug in an example in wls.Rd that triggers warning in R-Devel.
* Add Id, expCov, and expMean in create.mxModel().

Release 1.2.5 (Nov 29, 2020)
====================================
* Release to CRAN.
* Fix a bug in create.vechsR() with S0 argument to ensure that diagonals are 0.
* Fix a bug in create.Tau2() with the RE.type = "User" argument that incorrectly fixes all fixed parameters at 0.
* Add subset.rows argument in osmasem().
* Rename subset argument in osmasem() to subset.variables.
* Add an argument extraTries in rerun().
* Add dataset Mathieu15.
* Add calEffSizes().
* Add create.modMatrix(). 
* Rewrite Cor2DataFrame() to automatically append additional variables.

Release 1.2.4 (Jun 13, 2020)
====================================
* Release to CRAN.
* Add CFI and TLI in summary.osmasem(). 
* Rename the "Saturated" argument to "fitIndices" in summary.osmasem().
* Add "@" in the syntax of as.mxMatrix().
* Add as.mxAlgebra().
* Add as.symMatrix().
* Add an argument Jmatrix in osmasem().
* Add an argument ngroups in lavaan2RAM().
* Change the variable names in Becker92 and Becker94 so that they work better with lavaan2RAM().

Release 1.2.3.1 (Dec 8, 2019)
====================================
* Release to CRAN.
* Change class(x) == "foo" to inherits(x, "foo") to avoid errors in R 4.0.0.

Release 1.2.3 (Oct 10, 2019)
====================================
* Release to CRAN.
* Fix a minor bug in lavaan2RAM() that arranges variables according to x1, x10, x2, ...
* Improve tssem1FEM() in handling missing variables.
* Replace bootstrap vcov() with OpenMx's vcov() when there are constraints in wls().
* Add a subset argument in osmasem().
* Add a plotting method for osmasem object.
* Add metaFIML().
* Add an argument autofixtau2 for tssem1REM and osmasem objects in rerun().
* Add an argument RAM=NULL in wls(), tssem2(), and uniR2mx().
* Add dataset Kalaian96.
* Add robust argument in vcov.meta(), vcov.meta3X(), vcov.osmasem(), vcov.tssem1REM(), summary.meta(), summary.meta3X(), summary.osmasem(), and summary.tssem1REM().

Release 1.2.2 (May 29, 2019)
====================================
* Release to CRAN.
* Add dataset Stadler15.
* Add create.mxModel()

Release 1.2.0 (Oct 18, 2018)
====================================
* Release to CRAN.
* Add .onLoad()
* Add osmasem() and its related functions.
* Add an argument type=c("tssem", "osmasem") in pattern.na().
* Add datasets Gnambs18 and vanderPol17.

Release 1.1.0 (May 09, 2018)
====================================
* Release to CRAN.
* Add checkRAM().
* Add list.output argument in smdMTS() and smdMES().
* Add dataset Gleser94.
* Add dataset Scalco17.
* Modify plotting methods for lavaan models and wls objects.
* Fix a bug in list2matrix() for 2x2 matrices when the diag=FALSE.
* Fix a minor bug in asyCov() with acov="weighted" or "unweighted" when the cor.analysis=TRUE with covariance matrices as the input.

Release 1.0.0 (January 12, 2018)
====================================
* Release to CRAN.
* Add lavaan in import.
* Change the defaults of tssem1():
  		  From method="FEM" to method="REM",
  		  From RE.type="Symm" to RE.type="Diag", and
  		  From acov="individual" to acov="weighted."
* Add tssemParaVar() to estimate heterogeneity of parameter estimates in tssem approach.
* Add basic tests.
* Add fixed.x=FALSE automatically in uniR2lavaan().
* Fix coef.wls() so that it always follows the order of the parameter estimates in vcov.wls()
* Rename bootMASEM functions to bootuniR functions.
* Change the variable names in the output of asyCov().

Release 0.9.16 (September 29, 2017)
====================================
* Release to CRAN.
* Add bootMASEM functions.
* Rename the argument my.df to Cov in the tssem1 family.
* Rename the argument my.df to Cor in uniR1().
* Rename the argument asyCov to aCov in the wls() and tssem2().
* Rename the argument check.asyCov to check.aCov in is.pd().  
* Revise meta2semPlot() so that it generates nicer variable labels.
* Change the variable names in Becker94 so that they work better with lavaan2RAM().
* Revise lavaan2RAM() to work with mean structure.
* Revise as.mxMatrix() to work with definition variables.

Release 0.9.14 (May 18, 2017)
====================================
* Release to CRAN.
* Fix a minor bug reported by GerardCY in meta() when lm() fails to get the starting values for coef.constraints. 
* Add rCor(), rCorPop(), rCorSam(), summary.CorPop(), and print.summary.CorPop. 
* Fix a minor bug reported by John Ma when there are more than 120 variables in tssem1FEM().
* Revise the formula in calculating the standard error in uniR1() by using the second improved approximation of Schmidt and Hunter (2015, p. 101)
* Add smdMTS() and smdMES() to calculate standardized mean differences for multiple treatment studies and for multiple end-point studies.

Release 0.9.12 (January 23, 2017)
====================================
* Release to CRAN.
* Explicitly state "SLSQP" as the optimizer in metaSEM as "CSOLNP" is the default optimizer in OpenMx 2.7.4. 
* Revise and rename impliedSigma() to impliedR().
* Fix a bug in uniR2mx() that does not work for models with latent variables.
* Add dataset Cooke16.
* Change solve() to chol2inv(chol()) and ginv() for matrix inversion.
* Add dataset Boer16.
* Add uniR1(), uniR2mx(), and uniR2lavaan().
* Modify a variable name in Becker09.
* Minor changes on how errors are handled in the main functions. 
* Add datasets Nohe15A1 and Nohe15A2
* Modify some variable names in Hunter83

Release 0.9.10 (August 18, 2016)
====================================
* Release to CRAN.
* Add impliedSigma() to calculate the model implied correlation or covariance matrix based on the RAM model
* Add VarCorr() to extract variance component matrix for the class meta objects.
* Fix a minor bug in summary.tssem1FEM that throws an error when tT is NA.
* Add a check on the lengths of df and n in tssem1FEM() and tssem1REM().
* Change mxFitFunctionAlgebra() to mxFitFunctionMultigroup() in tssem1FEM().
* Fix a minor bug in lavaan2RAM() that does not honor the std.lv argument.
* Modify the vignette and added examples.
* Change the definition of is.pd() from Matrix::nearPD to MASS::mvrnorm().

Release 0.9.8 (April 16, 2016)
====================================
* Release to CRAN
* Return NA rather than error in calculating 95% CI of RMSEA when there are errors in summary.wls(). 
* Fix a minor bug in meta() that does not give a starting value on the last preditcor with intercept.constraints=0
* Add lavaan2RAM() to convert lavaan models to RAM models
* Add RMSEA 95% CI for wls and tssem1FEM classes

Release 0.9.6 (October 31, 2015)
====================================
* Release to CRAN
* Add several packages in NAMESPACE to satisfy R developmental version's (3.3.0) requirements
* Add ellipse package in imports
* Fix a bug reported by S.-F. Cheung in rerun() for tssem1FEM objects
* Add acov argument in asyCov() and tssem1()
* Modify wls(diag.constraints=FALSE) so that "mediators" are properly handled 
* Add datasets Norton13 and Roorda11
* Change default model.name in tssem1REM(), tssem1FEM() and wls() to avoid potential error code with OpenMx 2.2.6
* Add RE.constraints argument and "User" option in RE.type argument in tssem1() and tssem1REM()

Release 0.9.4 (June 7, 2015)
====================================
* Modify to fit the CRAN policies
* Remove deltaMethod()
* Revert the RE.lbound from NA to 1e10 in meta() and meta3()
* Change the default "Optimality tolerance" to "6.3e-14" in OpenMx
* Change the default "Gradient iterations" to "2" in OpenMx
* Change the default arguments in rerun()

Release 0.9.3-2 (May 25, 2015)
====================================
* Change the RE.lbound from 1e10 to NA in meta() and meta3()
* Add trunction (0 and 1) in calculating R2 in summary.meta()
* Change the default optimizer from "NPSOL" to "SLSQP" (default in OpenMx)
* Change the default "Gradient algorithm" to "central" in OpenMx
* Add metaSEM.Rnw in vignettes

Release 0.9.3-1 (May 12, 2015)
====================================
* Correct a typo in the example of Aloe14 (reported by Fred Li)
* Modify to fit R-3.2.0

Release 0.9-2 (Jan 23, 2015)
====================================
* Add dataset Cheung00
* Add check.asyCov and cor.analysis arguments in is.pd() 

Release 0.9-1 (Dec 22, 2014)
====================================
* Add citation in "Frontiers in Psychology"
* Release after the official release of OpenMx 2.0
* Remove dataset Craft03 which is duplicated of Becker09

Release 0.9-0 (Nov 16, 2014)
====================================
* Add semPlot package in Suggests and meta2semPlot() to convert wls objects to semPlotModel objects
* Fix a bug in tssem1REM() that may throw away correlation coefficients when there are missing correlations (not missing variables)
* Add byrow argument in vec2symMat()
* Add datasets Aloe14 and Becker09
* Remove ellipse package in the dependence
* Add silent=TRUE and run=TRUE arguments in most functions
* Fix the package to work in OpenMx 2.0
* Use labels rather than Amatrix and Smatrix in wls() and tssem2()
* Fix a bug in create.mxMatrix() to respect byrow=TRUE for "Symm" and "Stand" types 

Release 0.8-5 (Jun 6, 2014)
====================================
* Change "." to "_" in datasets Jaramillo05 and wvs94a, and indirectEffect.R
* Add pattern.n() to display the accumulative sample sizes of the matrices
* Add a function deltaMethod() to calculate approximate sampling variance or covariance matrix
* Add methods for coef.MxRAMModel and vcov.MxRAMModel
* Modify the formula to calculate SRMR in summary.tssem1FEM()
* Add xlim and ylim arguments in plot.meta()
* Add datasets wvs94b and BCG
* Rename the dataset wvs94 to wvs94a
* Revise is.pd() so that it returns NA rather than throws an error when there are NA in the input matrices
* Add byrow argument in create.mxMatrix()
* Add pattern.na() to display the pattern of missing data in TSSEM
* Fix a bug in asyCov() returning an error in diffferent dimensions when the input is a list of 2x2 matrices
* Fix a bug reported by Andre Plamondon that tssem1FEM() returns an error when acovS is a scalar 

Release 0.8-4 (May 1, 2013)
====================================
* Fix an issue in diag() that breaks the metaSEM in R-3.0.0
* Add Diag() that preserves the same functions as diag() prior to R-3.0.0
* Add dataset Craft03

Release 0.8-2 (Dec 28, 2012)
====================================
* Finalize features for release

Release 0.8-1 (Nov 28, 2012)
====================================
* Add automatic constraint on the diagonals when diag.constraints=FALSE argument in wls() for cor.analysis=TRUE. It is usually not necessary to set diag.constraints=TRUE unless there are mediators.
* Add meta3X() to handle missing covariates with FIML for 3-level meta-analysis
* Fix the issue that there are missing levels in "cluster" in meta3() (and reml3()) reported by David Stanley
* Modify meta(), reml(), meta3() and reml3() functions so that intercept.constraints, coef.constraints, RE.constraints, RE2.constraints and RE3.constraints may accept a scalar or vector as inputs
* Add mx.algebras argument in wls() and tssem2()
* Add datasets Hunter83 and Jaramillo05
* Change the default in as.mxMatrix. If 'x' is not a matrix, as.matrix(x) is applied on 'x' instead of throwing an error.

Release 0.8-0 (Aug 30, 2012)
====================================
* Finalize features for release

Release 0.7-1 (Aug 20, 2012)
====================================
* Remove argument RE_diag and added RE.type in tssem1() and tssem1REM()
* Remove standard error, z value and p values from summary() when intervals.type="LB"
* Add rerun() to refit models
* Rename all mx fitted objects to mx.fit
* Fix a bug in .minus2LL(), tssem1FEM() and summary.tssem1FEM in calculating the df of the independence model when cor.analysis=FALSE
* Fix a bug in tssem1REM() that my.df is a list of covariance matrices when cor.analysis=TRUE
* Fix a bug in calculating acov (solve()) in asyCov(), indirectEffect, tssem1.FE()
* Add indirectEffect() to calculate the asymptotic covariance matrix of indirect and direct effects
* Add create.Fmatrix() and create.mxMatrix() to create mxMatrix objects
* Add df.adjustment argument in summary.wls()
* Rewrite wls() using RAM specification (A, S and F matrices)
* Add diag.constraints argument in wls() to constrain diagonals as 1
* Improve efficiency by removing the call on installed.packages() and summary() from OpenMx
* Fix a bug in vec2symMat() reported by Yonghao Lim
* Add meta3() and reml3() for 3-level meta-analysis
* Add datasets Becker92, Becker94, Cooper03, Mak09, issp05 and Bornmann07
* Add I2 and R2 arguments to calculate heterogeneity and explained variance in meta()
* Rewrite meta() so that the predictors are treated as design matrix rather than as variables
* Revise meta() to replace NA with 1e5 in v object
* Rename coeff.constraints argument to coef.constraints in meta()
* Add homoStat argument in summary.meta()
* Rename tssem1FE() and tssem1RE() to tssem1FEM() and tssem1REM()

Release 0.7-0 (Nov 6, 2011)
====================================
* Fix a bug in meta() that does not include RE.lbound in Tau matrix when RE.lbound is a matrix
* Revise tssemFE() to handle incomplete data in the first group
* Add tssem1RE() for random-effects TSSEM
* Rename tssem1() to tssem1FE() for fixed-effects TSSEM
* Modify tssem1() to be a wrapper for tssem1FE() and tssem1RE()
* Modify tssem2() to fit both fixed- and random-effects structural models from tssem1()
* Add "cluster" argument in tssem1.FE() and some methods for it
* Add "select" argument in coef.meta() and vcov.meta()
* Fix a bug by removing the characters in matrix column when OpenMx 1.1 is used 
* Fix a bug in asyCov() that uses only the first sample size

Release 0.6-0 (May 24, 2011)
====================================
* Add a dataset WVS94
* Change dropNA argument from TRUE to FALSE in asyCov()
* Add data argument in meta() and reml()
* Fix a bug by addinng dimnames in M matrix in meta() when OpenMx 1.1 is used
* Fix a bug in summary.meta(), summary.reml() and summary.wls() when OpenMx 1.1 is used
* Add plot() method for multivariate meta-analysis
* Add the ellipse package on "Imports" in the DESCRIPTION
* Add the metafor package on "Suggests" in the DESCRIPTION

Release 0.5-4 (March 19, 2011)
====================================
* Fix a bug in tssem1() when OpenMx 1.0.6 is used
* Fix a bug in tssem1() which ignores start.values argument

Release 0.5-3 (February 20, 2011)
====================================
* Add model.name argument in tssem1(), tssem2(), wls(), meta() and reml()
* Add anova() method for wls, meta and reml objects
* Add no. of studies and no. of observed statistics in reml object (Ad-hoc) 
* Use Tau2 to represent variance component in both meta() and reml()
* Fix a bug in meta() that exists when no. of predictors is larger than 1

Release 0.5-2 (December 11, 2010)
====================================
* Add name argument in as.mxMatrix()
* Add reml() to estimate variance components with REML

Release 0.5-1 (October 10, 2010)
====================================
* Add bdiagMat() and bdiagRep() functions to create block diagonal matrix
* Fix a bug in meta() that ignored RE.lbound when RE.constraints was used
* Fix a bug in readFullMat() that converts data into data.frames rather than matrices

Release 0.5-0 (October 2, 2010)
====================================
* Add vcov() methods for tssem1, wls objects
* Add coef() methods for tssem1, wls and meta objects

Release 0.3-9 (September 24, 2010)
====================================
* Add a vignette

Release 0.3-8 (September 20, 2010)
====================================
* Add suppressWarnings = TRUE in meta(), tssem1(), tssem2(), and wls()
* Add OpenMx.status[[1]] into summary method

Release 0.3-7 (September 19, 2010)
====================================
* Add vec2symMat(), matrix2bdiag() and homoStat() functions
* Move startValues(), minus2LL() and indepwlsChisq() to hidden functions
* Update datasets on Cheung09 and Digman97

Release 0.3-6 (September 16, 2010)
====================================
* Rename the package to "Meta-Analysis using Structural Equation Modeling"
* Add vcov() method for meta object

Release 0.3-5 (September 14, 2010)
====================================
* Fix a few bugs

Release 0.3-4 (September 12, 2010)
====================================
* Add summary.meta() for meta objects

Release 0.3-3 (September 8, 2010)
====================================
* Add new arguments in asyCov()
* Add meta() for univariate and multivariate meta-analysis
* Add as.mxMatrix() to convert matrices to mxMatrices

Release 0.3-2 (August 29, 2010)
====================================
* Add indepWLSChisq() to calculate the chi-square statistic for an independent model based on WLS
* Add minus2LL() to calculate the chi-square statistic for independent and saturated models based on ML
* Add summary.wls() and print.summary.wls() for wls objects
* Add summary.tssem1() and print.summary.tssem1() for tssem1 objects

Release 0.3-1 (August 09, 2010)
====================================
* First alpha release to R-forge
* Add basic functions for TSSEM
