aboutsummaryrefslogtreecommitdiff
path: root/R/mkinfit.R
AgeCommit message (Collapse)AuthorFilesLines
2015-06-21Do the t-test for untransformed parametersv0.9-36Johannes Ranke1-18/+71
2015-06-20Low-level generation of compiled modelsJohannes Ranke1-3/+3
As it is unclear if and when ccSolve will be published on CRAN, the generation, compilation and use of the C version of the system of differential equations was developed for mkin, inspired and guided by the code from the ccSolve package. Many thanks again to Karline Soetaert for all of her work on this and other R packages. Now all model types, including the Hockey-Stick model for the parent compund and the IORE model for parent and/or metabolites can be compiled.
2015-04-16Update copyright header, improve documentation of solution_typeJohannes Ranke1-12/+17
2015-04-16Check compiled mkinmod versions in examples and testsJohannes Ranke1-0/+3
Tests all pass (Linux)
2015-02-20Add termination message to mkinfit, prepare releaseJohannes Ranke1-1/+9
2015-02-20Fix printing the summary if iterations are not availableJohannes Ranke1-1/+4
This was broken for using method.modFit = "SANN"
2014-11-12IORE working for parent and metabolites, introduce mkinsubJohannes Ranke1-2/+7
2014-11-12Merge branch 'master' into ioreJohannes Ranke1-48/+127
2014-11-12Fix a typo in the list of possible argumentsJohannes Ranke1-1/+1
Thanks to Michael Brauer for the hint
2014-10-22Always include 0 on y axis when plotting during the fitJohannes Ranke1-1/+1
2014-10-15Switch to using the Port algorithm per defaultJohannes Ranke1-1/+1
2014-10-14Further formatting improvement for differential equationsJohannes Ranke1-1/+1
2014-10-06Avoid a warning produced by initial fits in gmkinJohannes Ranke1-7/+9
2014-08-20Fix mkinfit for a special situationJohannes Ranke1-7/+7
See NEWS.md: When the parent was not the (only) variable with the highest value out of all variables in the observed data, it could happen that the initial value (state.ini) was not initialised.
2014-08-19Improve formatting of differential equations in outputJohannes Ranke1-1/+1
Rebuild of FOCUS_Z vignette with improved formatting
2014-07-24Bump version, better default for state.iniJohannes Ranke1-2/+8
2014-07-21Bugfix, model shorthand, state.ini[[1]] from observed dataJohannes Ranke1-3/+24
- The bug occurred when using transform_rates=FALSE for FOMC, DFOP or HS - Make it possible to use mkinfit("SFO", ...) - Take initial mean value at time zero for the variable with the highest value in the observed data - Update of vignette/FOCUS_L - Improve the Makefile to build single vignettes
2014-07-17Return complete list of initial states after fittingJohannes Ranke1-1/+4
This is useful for specifying state.ini in a subsequent call to mkinfit
2014-07-17Handle non-convergence and maximum number of iterationsJohannes Ranke1-35/+82
For details see NEWS.md
2014-07-14Merge bugfix branch 'master' into ioreJohannes Ranke1-19/+17
Add IORE support to mkinerrmin
2014-07-14Bugfix release version 0.9-31Johannes Ranke1-19/+17
2014-07-12First working fits with IORE modelJohannes Ranke1-1/+3
2014-07-02Do not introduce formation fractions for singular pathwaysJohannes Ranke1-23/+16
- As a consequence, we do not need implicit formation fractions in mkinfit
2014-07-02Warning -> message, and only if initial parameter was not 1Johannes Ranke1-2/+4
2014-07-02Set implicit formation fraction to 1 even if set otherwise in parms.iniJohannes Ranke1-6/+8
2014-07-02Move handling of implicitly fixed formation fractions to mkinfitJohannes Ranke1-13/+11
2014-07-02Support formation fractions without sink pathway, updatesJohannes Ranke1-50/+113
2014-07-01One more source code commentJohannes Ranke1-0/+2
2014-06-27Reorganize the summary in a more logical orderJohannes Ranke1-19/+19
2014-05-20Step back regarding confidence intervals of formation fractionsJohannes Ranke1-3/+15
- Don't give confidence intervals for formation fractions that belong to a set, only for single formation fractions - Update the vignette with history and background - Correct 'isotropic' into 'isometric' in the ilr documentation, the README and the comments in transform_odeparms - Candidate for release to CRAN
2014-05-09Set upper and lower boundx for untransformed fractionsJohannes Ranke1-3/+16
2014-04-30Fix of bug introduced with t-testJohannes Ranke1-1/+1
2014-04-30Restore t-test (1- and 2-sided) for fitted parametersJohannes Ranke1-3/+6
2014-04-25McCall 2,4,5-T dataset, lower bound for transform_rates=FALSEJohannes Ranke1-2/+11
2014-04-22Possibility to fit without parameter transformationJohannes Ranke1-8/+26
2014-02-27Format of gmkin workspace files, copyright, updatesJohannes Ranke1-0/+1
- Change the format of the workspace files of gmkin so they can be distributed in the mkin package - Add gmkin workspace datasets FOCUS_2006_gmkin and FOCUS_2006_Z_gmkin - Restrict the unit test with the Schaefer data to parent and primary metabolites as formation fraction and DT50 for A2 are higly correlated and passing the test is platform dependent. For example, the test fails in 1 out of 14 platforms on CRAN as of today - Add Eurofins Regulatory AG copyright notices - Import FME and deSolve instead of depending on them to have cleaner startup
2014-01-22Merge branch 'master' of https://github.com/jranke/mkinJohannes Ranke1-3/+5
Conflicts: DESCRIPTION R/mkinfit.R
2014-01-17- Use mkinmod$spec instead of mkinmod$map wherever possible, in order to ↵Johannes Ranke1-4/+4
make it possible to to modify mkinmod$map when the mapping of internal to observed variables should be changed. Note that such modifications will confuse plot.mkinfit and summary.mkinfit. Setting the class attribute to modFit makes summary.modFit useable in such a case
2013-12-04Make it possible to use different ode solvers=1-3/+5
2013-11-13Squashed commit of the following:jranke1-0/+2
commit d8dbf2ad866fb9d34cd1100000b9c116219ecef6 Author: Johannes Ranke <jranke@uni-bremen.de> Date: Wed Nov 13 22:09:41 2013 +0100 Add possibility to control more mkinfit arguments from the GUI commit 8f1bae2142b37a0ff6b8989b2d1569686937f68e Author: Johannes Ranke <jranke@uni-bremen.de> Date: Wed Nov 13 20:30:11 2013 +0100 Add initial weighting choice to GUI commit c99b5c298713a7c14e8ab5604c68613d0b7af27a Author: Johannes Ranke <jranke@uni-bremen.de> Date: Wed Nov 13 16:25:31 2013 +0100 Update of TODO from git bash on Windows commit 8bf2a6f8f6bd752433f06b26b5da334e958f5166 Author: Johannes Ranke <jranke@uni-bremen.de> Date: Wed Nov 13 16:00:01 2013 +0100 Update of README using TortoiseGit git-svn-id: svn+ssh://svn.r-forge.r-project.org/svnroot/kinfit/pkg/mkin@162 edb9625f-4e0d-4859-8d74-9fd3b1da38cb
2013-11-05- Use an increasing small numeric value instead of random numbers for ↵jranke1-1/+5
generating different default values for kinetic rate constants to increase transparency and reproducibility. This also works nicely. - Remove an outdated comment from the test script. git-svn-id: svn+ssh://svn.r-forge.r-project.org/svnroot/kinfit/pkg/mkin@137 edb9625f-4e0d-4859-8d74-9fd3b1da38cb
2013-11-04Add small random numbers to default initial values for rate constants to avoidjranke1-2/+2
numerical artefacts occurring in eigenvalue based solutions. git-svn-id: svn+ssh://svn.r-forge.r-project.org/svnroot/kinfit/pkg/mkin@135 edb9625f-4e0d-4859-8d74-9fd3b1da38cb
2013-11-04Bugfix for a recently introduced breakage of fixed_initialsjranke1-1/+4
git-svn-id: svn+ssh://svn.r-forge.r-project.org/svnroot/kinfit/pkg/mkin@132 edb9625f-4e0d-4859-8d74-9fd3b1da38cb
2013-10-30- Backtransform fixed ODE parameters for the summary (compare ChangeLog)jranke1-2/+2
- Reformatted ChangeLog - Version bump as mkin 0.9-22 is on CRAN git-svn-id: svn+ssh://svn.r-forge.r-project.org/svnroot/kinfit/pkg/mkin@129 edb9625f-4e0d-4859-8d74-9fd3b1da38cb
2013-10-29- Just one little formatting change in mkinfit.Rjranke1-1/+2
- Some very basic fitting now works in the experimental GUI with the new widget layout git-svn-id: svn+ssh://svn.r-forge.r-project.org/svnroot/kinfit/pkg/mkin@128 edb9625f-4e0d-4859-8d74-9fd3b1da38cb
2013-10-17- Fix reweighting for the case of many metabolitesjranke1-1/+1
git-svn-id: svn+ssh://svn.r-forge.r-project.org/svnroot/kinfit/pkg/mkin@118 edb9625f-4e0d-4859-8d74-9fd3b1da38cb
2013-10-16- More consistent output regarding optimised and fixed parameters (see ↵jranke1-4/+9
ChangeLog) - Switch from gcanvas to gsvg - Start setting up the fits in the GUI - add files for testing gWidgetsWWW2 functionality git-svn-id: svn+ssh://svn.r-forge.r-project.org/svnroot/kinfit/pkg/mkin@115 edb9625f-4e0d-4859-8d74-9fd3b1da38cb
2013-10-10- IRLS is implementedjranke1-14/+58
git-svn-id: svn+ssh://svn.r-forge.r-project.org/svnroot/kinfit/pkg/mkin@109 edb9625f-4e0d-4859-8d74-9fd3b1da38cb
2013-10-10- Added a ChangeLogjranke1-16/+33
- Do not use time zero values of 0 for chi2 error level calculations (see changelog) - Show weighting method in summary - Correct the output of the data in the case of manual weights - Some reformatting in mkinfit.R - GUI: First attempt at representing a fit in the GUI git-svn-id: svn+ssh://svn.r-forge.r-project.org/svnroot/kinfit/pkg/mkin@108 edb9625f-4e0d-4859-8d74-9fd3b1da38cb
2013-06-18Pending check-in of stuff started at east one month ago:jranke1-0/+3
- avoid returning factors in the data frame in mkin_wide_to_long - stop returning useless results with eigenvalue based solutions when observed absolute values are all very small - better test case for schaefer97 data making use of stepwise fitting - but the solution obtained in one step is better... git-svn-id: svn+ssh://svn.r-forge.r-project.org/svnroot/kinfit/pkg/mkin@83 edb9625f-4e0d-4859-8d74-9fd3b1da38cb

Contact - Imprint