12590c3 | Joep Vanlier | 08 December 2015, 13:42:57 UTC | Added the option to include input contributions in the DERIVED variables. | 08 December 2015, 13:42:57 UTC |
7bdc64c | Joep Vanlier | 08 December 2015, 13:05:35 UTC | Resolved corner case bug regarding sparse solver. | 08 December 2015, 13:05:35 UTC |
17e800c | Joep Vanlier | 08 December 2015, 10:40:23 UTC | removed negfluxes check for example | 08 December 2015, 10:40:23 UTC |
455b4a6 | Fabian Konrath | 07 December 2015, 14:27:59 UTC | merge | 07 December 2015, 14:27:59 UTC |
150996d | Fabian Konrath | 07 December 2015, 14:00:26 UTC | Added option ar.config.useFitErrorMatrix which allows to fit models simultaneously to different types of data (fiterrors=1 vs fiterrors=-1). To use this option ar.config.useFitErrorMatrix has to be set to true and the different fiterrros have to be specified in ar.config.fiterrors_matrix: ar.config.fiterrors_matrix(model index, data index) = 1 | 0 | -1 For correct plotting, ploterrors can be specified in ar.config.ploterrors_matrix: ar.config.ploterrors_matrix(model index, data index) = 1 | 0 | -1 The likelihood is calculated as follows: -2*LL=2*ar.ndata_err*log(sqrt(2*pi))+ar.chi2+ar.chi2err where ar.ndata_err is the sum of all data points of data sets with fiterrors=1, ar.chi2 is the sum of all chi2 values (of all data sets), ar.chi2err is the sum of all chi2errs of data sets with fiterrors=1. | 07 December 2015, 14:00:26 UTC |
6c80c1f | Andreas Raue | 05 December 2015, 20:50:48 UTC | bug fix and nicer font | 05 December 2015, 20:50:48 UTC |
ceaf262 | Andreas Raue | 05 December 2015, 20:39:07 UTC | Improves in report function | 05 December 2015, 20:39:07 UTC |
c537115 | Andreas Raue | 04 December 2015, 14:41:18 UTC | Some improvements in the latex report | 04 December 2015, 14:41:18 UTC |
55887cd | Joep Vanlier | 03 December 2015, 14:19:16 UTC | Forgot the actual function :) | 03 December 2015, 14:19:16 UTC |
5eaadf2 | Joep Vanlier | 03 December 2015, 14:17:29 UTC | merge | 03 December 2015, 14:17:29 UTC |
b57dce8 | Joep Vanlier | 03 December 2015, 14:16:51 UTC | Moved reserved word checking to load time, rather than compile time (faster and clearer error messages). | 03 December 2015, 14:16:51 UTC |
011abb1 | Clemens Blank | 03 December 2015, 13:02:17 UTC | Added the d2d-presenter to the d2d repo. | 03 December 2015, 13:02:17 UTC |
da84027 | Clemens Blank | 03 December 2015, 12:47:54 UTC | test deleted online with Bitbucket | 03 December 2015, 12:47:54 UTC |
d7535a3 | Clemens Blank | 03 December 2015, 12:47:22 UTC | test created online with Bitbucket | 03 December 2015, 12:47:22 UTC |
be22b4f | Andreas Raue | 02 December 2015, 10:07:53 UTC | Merge | 02 December 2015, 10:07:53 UTC |
83d1280 | Andreas Raue | 02 December 2015, 09:56:12 UTC | added support for new location of pdflatex on mac el capitan | 02 December 2015, 09:56:12 UTC |
ce7face | Andreas Raue | 01 December 2015, 17:29:00 UTC | utility functions | 01 December 2015, 17:29:00 UTC |
86e78b4 | Andreas Raue | 01 December 2015, 16:01:47 UTC | some update in the report function | 01 December 2015, 16:01:47 UTC |
ce1e71e | Andreas Raue | 01 December 2015, 13:17:50 UTC | I seperated out the subs function, to clean up code and make it more modular. This might lead to some smaller bugs in some of the less used function but should be mostly okay as far as I can tell right now. | 01 December 2015, 13:17:50 UTC |
eb41c0a | Andreas Raue | 30 November 2015, 19:03:29 UTC | Changed default behaviour of enforced positive values to be 0, i.e. value are not forced >0. I think this is the better default. Some model actually have states that can become negative and this seems to prevent some CV_ERR_TOL errors as well. | 30 November 2015, 19:03:29 UTC |
bfcab22 | Joep Vanlier | 30 November 2015, 13:23:39 UTC | Added some arPlotY options. | 30 November 2015, 13:23:39 UTC |
d5e5ab4 | Joep Vanlier | 27 November 2015, 16:37:18 UTC | Added optional optimization during fitting. Only compute sensi's when requested (some solvers also use regular function evals). | 27 November 2015, 16:37:18 UTC |
6c4422b | Joep Vanlier | 26 November 2015, 17:37:37 UTC | Improved error messages. | 26 November 2015, 17:37:37 UTC |
9f0298e | Joep Vanlier | 26 November 2015, 17:21:59 UTC | Changed mex file output to be more in agreement with lsqnonlin. | 26 November 2015, 17:21:59 UTC |
f256962 | Joep Vanlier | 26 November 2015, 16:11:04 UTC | Added NL2SOL to arFit. Give it a spin (ar.config.optimizer = 8). First call it will compile, but this shouldn't take too long. | 26 November 2015, 16:11:04 UTC |
7157f78 | Joep Vanlier | 26 November 2015, 15:17:19 UTC | NL2SOL files from PORT library (commit 2) | 26 November 2015, 15:17:19 UTC |
93828e5 | Joep Vanlier | 26 November 2015, 15:12:59 UTC | New optimization routine (first commit): NL2SOL adaptive non-linear trust region based optimizer from the PORT library. | 26 November 2015, 15:12:59 UTC |
00972e7 | Bernhard Steiert | 25 November 2015, 09:42:55 UTC | Small fix | 25 November 2015, 09:42:55 UTC |
1aea536 | Bernhard Steiert | 24 November 2015, 10:57:15 UTC | Bugfix: arCheckCache now works with older workspaces that have less ODE solver options | 24 November 2015, 10:57:15 UTC |
b4cabac | Clemens Kreutz | 23 November 2015, 12:07:25 UTC | New example (without data): Huang Map-Kinase | 23 November 2015, 12:07:25 UTC |
78b9def | Clemens Kreutz | 23 November 2015, 12:04:52 UTC | New example (without data): Restif Vaccination | 23 November 2015, 12:04:52 UTC |
14daa57 | Clemens Kreutz | 23 November 2015, 12:04:33 UTC | New example (without data): Steckmann Fibrillogenesis | 23 November 2015, 12:04:33 UTC |
c322a3a | Clemens Kreutz | 23 November 2015, 12:04:03 UTC | New example (without data): Kholodenko EGFR | 23 November 2015, 12:04:03 UTC |
48fce71 | Clemens Kreutz | 23 November 2015, 12:02:26 UTC | New example (without data): Fribourg Immune Response H1N1 | 23 November 2015, 12:02:26 UTC |
bf34e15 | Clemens Kreutz | 23 November 2015, 12:01:43 UTC | New example (without data): Elowitz Repressilator | 23 November 2015, 12:01:43 UTC |
d271f08 | Clemens Kreutz | 23 November 2015, 12:01:04 UTC | New example (without data): Dalla Man Insulin | 23 November 2015, 12:01:04 UTC |
394cebf | Clemens Kreutz | 23 November 2015, 11:59:06 UTC | Another improvement: "I" and "E" are treated as variables instead of imaginary i and exp(1) | 23 November 2015, 11:59:06 UTC |
5e1c052 | Clemens Kreutz | 23 November 2015, 10:58:49 UTC | Improved SMBL import (tested with 10 models from BioModels Database) incl. related adaption in loading of data and models. | 23 November 2015, 10:58:49 UTC |
65783a5 | Joep Vanlier | 23 November 2015, 10:12:16 UTC | Check whether user has required licenses on arInit | 23 November 2015, 10:12:16 UTC |
99cf045 | Joep Vanlier | 20 November 2015, 15:00:08 UTC | Added caching of all solver related variables to check against when deciding on whether to resimulate sensitivities or not. | 20 November 2015, 15:00:08 UTC |
c6fcbae | Joep Vanlier | 20 November 2015, 13:42:29 UTC | merge | 20 November 2015, 13:42:29 UTC |
5da3d6f | Helge Hass | 20 November 2015, 13:50:29 UTC | better performance of prediction bands | 20 November 2015, 13:50:29 UTC |
8eb464f | Joep Vanlier | 20 November 2015, 13:42:11 UTC | bugfix | 20 November 2015, 13:42:11 UTC |
f984958 | Joep Vanlier | 20 November 2015, 13:36:19 UTC | Better fix for corner case | 20 November 2015, 13:36:19 UTC |
1cb4054 | Joep Vanlier | 20 November 2015, 13:26:39 UTC | Bugfix for corner case dynamics+sensi, dynamics+nosensi, nodynamics+sensi | 20 November 2015, 13:26:39 UTC |
18fb851 | Joep Vanlier | 20 November 2015, 13:07:42 UTC | merge | 20 November 2015, 13:07:42 UTC |
7fcbcae | Joep Vanlier | 20 November 2015, 13:07:27 UTC | Fix redundant sensitivity simulation | 20 November 2015, 13:07:27 UTC |
40cd3f2 | Bernhard Steiert | 20 November 2015, 09:21:03 UTC | Bugfix PPL calculation | 20 November 2015, 09:21:03 UTC |
6396328 | Joep Vanlier | 18 November 2015, 16:46:50 UTC | merge | 18 November 2015, 16:46:50 UTC |
0ef2741 | Joep Vanlier | 18 November 2015, 16:46:42 UTC | Fixed incorrect xLabel in plePlotMulti (would display as log10 despite being in linear space). | 18 November 2015, 16:46:42 UTC |
bd2e1bf | Bernhard Steiert | 18 November 2015, 13:20:20 UTC | Small fix in plotting of prediction bands | 18 November 2015, 13:20:20 UTC |
a409a13 | Joep Vanlier | 17 November 2015, 14:53:57 UTC | Better error message. | 17 November 2015, 14:53:57 UTC |
2d45d8c | Joep Vanlier | 17 November 2015, 14:27:59 UTC | Increase substitution cycles. | 17 November 2015, 14:27:59 UTC |
a7771d6 | Joep Vanlier | 17 November 2015, 14:16:13 UTC | Added SUBSTITUTIONS field to model and data, to allow for parameter substitutions (always specified before CONDITIONS). | 17 November 2015, 14:16:13 UTC |
4fec917 | Joep Vanlier | 16 November 2015, 17:39:25 UTC | Partial implementation of hierarchical substitutions (disabled for now). | 16 November 2015, 17:39:25 UTC |
28cd9ff | Joep Vanlier | 13 November 2015, 17:41:39 UTC | Bugfix. Replacing by ones sometimes leads to collisions which can be problematic when substituting them in certain equations. | 13 November 2015, 17:41:39 UTC |
77ff485 | Clemens Kreutz | 13 November 2015, 09:01:46 UTC | Forgot to uncomment some lines. | 13 November 2015, 09:01:46 UTC |
eb859a5 | Clemens Kreutz | 13 November 2015, 08:44:47 UTC | Bug fix in arLhsSampleSizeCalculation: missing function Runde() added. | 13 November 2015, 08:44:47 UTC |
5e67881 | Joep Vanlier | 12 November 2015, 18:13:09 UTC | Fixed tricky bug. Model would not resimulate when dynamic parameters were not being fitted, but were actually different from the ones that were simulated last. Now it checks against a reference list. | 12 November 2015, 18:13:09 UTC |
0eb7ff7 | Joep Vanlier | 12 November 2015, 13:15:27 UTC | remove globals in subfunctions | 12 November 2015, 13:15:27 UTC |
5b6620b | Joep Vanlier | 12 November 2015, 12:50:01 UTC | Pass matlab version along compilation chain to prevent race condition arising from the use of globals in threaded routines. | 12 November 2015, 12:50:01 UTC |
9fd399d | Joep Vanlier | 12 November 2015, 10:18:36 UTC | Summary stats for arLoadPars | 12 November 2015, 10:18:36 UTC |
0b03ed5 | Joep Vanlier | 11 November 2015, 17:51:25 UTC | Reverse accidental upload of edit in Swameye setup file. | 11 November 2015, 17:51:25 UTC |
c4b10e5 | Joep Vanlier | 11 November 2015, 17:49:22 UTC | Bugfix arFindInputs.m when the condition variables specify a numeric value of an input. | 11 November 2015, 17:49:22 UTC |
b42dd88 | Joep Vanlier | 10 November 2015, 14:32:20 UTC | Merge | 10 November 2015, 14:32:20 UTC |
fd55864 | Joep Vanlier | 10 November 2015, 14:32:11 UTC | Cached MATLAB version check (25% reduction in compilation time on my system) | 10 November 2015, 14:32:11 UTC |
3c1313d | Andreas Raue | 10 November 2015, 13:56:26 UTC | bug fix for cases where output was requested | 10 November 2015, 13:56:26 UTC |
cad3c62 | Clemens Kreutz | 10 November 2015, 13:49:08 UTC | Bug fix occuring in arLhsSampleSizeCalculation if called via arFitLHS(smallNumber) | 10 November 2015, 13:49:08 UTC |
f2c1476 | Joep Vanlier | 09 November 2015, 16:36:25 UTC | merge | 09 November 2015, 16:36:25 UTC |
161c8d5 | Joep Vanlier | 09 November 2015, 16:34:00 UTC | Added support for RANDOMs in arExport. Improved figure legend location arPlotY. Suppressed annoying output arPlotter. | 09 November 2015, 16:34:00 UTC |
8b9edf5 | Max Schelker | 09 November 2015, 15:23:10 UTC | minor | 09 November 2015, 15:23:10 UTC |
1f86a5f | Lorenz Adlung | 04 November 2015, 14:01:10 UTC | arMergeFits of chosen workspaces edited online with Bitbucket | 04 November 2015, 14:01:10 UTC |
ea59386 | Helge Hass | 03 November 2015, 14:09:15 UTC | minor fix | 03 November 2015, 14:09:15 UTC |
73856a5 | Max Schelker | 02 November 2015, 13:26:04 UTC | Merge | 02 November 2015, 13:26:04 UTC |
f32fc0c | Max Schelker | 02 November 2015, 13:25:46 UTC | Merge | 02 November 2015, 13:25:46 UTC |
5f3ff12 | Clemens Kreutz | 01 November 2015, 09:29:06 UTC | 1) arFitLHS now calls arLhsSampleSizeCalculation.m 2) Call arFitLHS(8,[],[],true) is now feasible. | 01 November 2015, 09:29:06 UTC |
2b88184 | Clemens Kreutz | 31 October 2015, 21:59:02 UTC | New function: Sample size calculation for LHS | 31 October 2015, 21:59:02 UTC |
adae663 | Clemens Kreutz | 31 October 2015, 20:59:43 UTC | Bug fix (error if field does not exist) | 31 October 2015, 20:59:43 UTC |
19369d2 | Clemens Kreutz | 31 October 2015, 20:53:15 UTC | Bug fix (error if field does not exist) | 31 October 2015, 20:53:15 UTC |
fcff228 | Max Schelker | 30 October 2015, 15:50:16 UTC | minor | 30 October 2015, 15:50:16 UTC |
2c64a18 | Clemens Kreutz | 30 October 2015, 14:13:49 UTC | Small change in the Setup of the DREAM Example. | 30 October 2015, 14:13:49 UTC |
18513ee | Clemens Kreutz | 30 October 2015, 14:11:15 UTC | New ple functionalities: 1) pleGlobals.mode=5 enforces that the threshold or the bounds are reached by the number of specified steps (samplesize). Usually the samplesize can be decreased compared to pleGlobals.mode=1. Can only be used via arPLEInit([],[],5). 2) ple with option "force" 3) Loading and saving: pleGlobals is also saved in workspace.mat if arSave is executed. It is used if the profile calculation saved in result.mat is not finished. This prevents unintended overwriting of old results if ple is aborted. | 30 October 2015, 14:11:15 UTC |
018854a | Joep Vanlier | 30 October 2015, 10:28:39 UTC | Merge | 30 October 2015, 10:28:39 UTC |
b5a8da4 | Joep Vanlier | 30 October 2015, 10:28:21 UTC | Added more descriptive error messages for incorrectly specified input functions. | 30 October 2015, 10:28:21 UTC |
6fbf3c1 | Bernhard Steiert | 29 October 2015, 13:59:29 UTC | Fix issue #49 | 29 October 2015, 13:59:29 UTC |
a812cb6 | Joep Vanlier | 27 October 2015, 14:39:47 UTC | minor | 27 October 2015, 14:39:47 UTC |
bbd69fa | Joep Vanlier | 27 October 2015, 14:37:25 UTC | bugfix | 27 October 2015, 14:37:25 UTC |
56d3edd | Joep Vanlier | 27 October 2015, 14:32:28 UTC | Added option to add description to data file input change. | 27 October 2015, 14:32:28 UTC |
21ded04 | Joep Vanlier | 27 October 2015, 09:45:29 UTC | merge | 27 October 2015, 09:45:29 UTC |
981aaac | Joep Vanlier | 27 October 2015, 09:45:11 UTC | Added sensitivity RHS contributions. Fitting compartment volumes should now be possible without problems. | 27 October 2015, 09:45:11 UTC |
512b6fd | Andreas Raue | 23 October 2015, 20:12:29 UTC | Remove all paper related files from Repository | 23 October 2015, 20:12:29 UTC |
8a915bd | Joep Vanlier | 23 October 2015, 17:20:28 UTC | Fixed minor bug in arPrint | 23 October 2015, 17:20:28 UTC |
64107d3 | Joep Vanlier | 23 October 2015, 16:39:31 UTC | Clemens's earlier push | 23 October 2015, 16:39:31 UTC |
dce4c21 | Joep Vanlier | 23 October 2015, 16:35:54 UTC | Reverted bad merge | 23 October 2015, 16:35:54 UTC |
5748b5b | Joep Vanlier | 23 October 2015, 15:34:56 UTC | Merge | 23 October 2015, 15:34:56 UTC |
19ee0b1 | Joep Vanlier | 23 October 2015, 15:34:44 UTC | Added warning & sensitivities for volume parameters | 23 October 2015, 15:34:44 UTC |
0fbbd2c | Clemens Kreutz | 23 October 2015, 13:41:13 UTC | Zusammenführen | 23 October 2015, 13:41:13 UTC |
2cdc8da | Clemens Kreutz | 23 October 2015, 13:15:09 UTC | New option for logging within optimization, new config: ar.config.logFitting, new sub-function in arFit.m: arLogFitDetailed. | 23 October 2015, 13:15:09 UTC |
48b343b | Joep Vanlier | 23 October 2015, 11:49:50 UTC | Fixed bug associated with new negative state detection (field may not exist when there is no data) | 23 October 2015, 11:49:50 UTC |
b88c1ca | Joep Vanlier | 23 October 2015, 11:44:25 UTC | Added free volume parameters to parameters list. To do: appropriate sensitivities | 23 October 2015, 11:44:25 UTC |