Forum
Tested V7.2 for model 2 fitting - now works fine - Printable Version

+- Forum (https://www.pcmodfit.co.uk/Forum)
+-- Forum: My Category (https://www.pcmodfit.co.uk/Forum/forumdisplay.php?fid=1)
+--- Forum: My Forum (https://www.pcmodfit.co.uk/Forum/forumdisplay.php?fid=2)
+--- Thread: Tested V7.2 for model 2 fitting - now works fine (/showthread.php?tid=6)



Tested V7.2 for model 2 fitting - now works fine - Notquitenewton - 05-04-2022

To the author and anyone else...
I've been messing around and trying out PK data sets for fitting! Just in case anyone wants to try out the corrected model 2 issue...attached are some data that I used for testing it. V7.1 and earlier versions were troublesome with these sets for model 2 but ok for model 9. Now works brilliantly with both models in V7.2 (thanks Graham!!). The Excel sheet attached shows the set up that I used...
Notquitenewton
Smile Smile


RE: Tested V7.2 for model 2 fitting - now works fine - ILikePK - 05-04-2022

To Notquitenewton
Thanks for the data sets. Just tried these out and yes, good result! Think I mentioned before; having the graphics shown in the spreadsheet is very smooth. The modelling of the supplied data sets runs so fast...generating the starting estimates automatically, fitting the data and graphs etc. only took a second or so. Great stuff!
ILikePK


RE: Tested V7.2 for model 2 fitting - now works fine - Administrator - 05-04-2022

Thanks to Notquitenewton and ILikePK for the data sets and testing out model 2 and for your +ve comments. I personally tested over 20 different profiles with various points added and removed and so far it's behaving very well.
One thing to note about modelling oral profiles can be simply put...as there are often several processes going on during an absorption phase, a single rate constant (ka) does not always define exactly what is going on for some data sets and this is where a Superposition approach may become preferable.
Lastly, as some data sets have very sparse data during the absorption phase (sometimes none) it may be necessary to utilise parameter constraints to avoid V and ka, as examples, shooting off into space with errors, orders of magnitude larger than the parameters themselves!!
Thanks again for your continued interest.
I note that the Forum is still fairly quiet at present but will hopefully pickup when more PK people join.
Administrator (Graham)