PdfFit2 vs PdfGui

188 views
Skip to first unread message

Fred marlton

unread,
Oct 26, 2017, 8:08:46 PM10/26/17
to diffpy-users
Hi All,

Is there any difference between doing refinements in PdfFit2 and PdfGui? (aside from the GUI)

I personally prefer using PdfFit2 and writing python scripts as it's similar to writing input files for Topas-Academic and great for batch fitting. But I found PdfGui to be really useful for getting started.

However, many times I have found a refinement to work better in PdfGui. (e.g. singular matrix errors less common or converges on a lower Rw)

Thanks,
Fred

Pavol Juhas

unread,
Oct 27, 2017, 6:34:16 PM10/27/17
to diffpy-users
On Thu, Oct 26, 2017 at 05:08:46PM -0700, Fred marlton wrote:
...
> Is there any difference between doing refinements in PdfFit2 and PdfGui?
> (aside from the GUI)

Hi Fred,

PDFfit2 is the fitting engine for PDFgui therefore it will
give the same results given a refinement problem defined in
equivalent way. Besides GUI and data organization, the main
extra features in PDFgui are space group symmetry functions
for expanding unit cell and generating symmetry-related
constraints. Although these can be specified in PDFfit2
scripts as well, there are no helper functions so it would
be quite tedious and error prone task.

> I personally prefer using PdfFit2 and writing python scripts as it's
> similar to writing input files for Topas-Academic and great for batch
> fitting. But I found PdfGui to be really useful for getting started.

If you need batch fitting beyond the predefined PDFgui macros,
perhaps the diffpy.pdfgui.tui module is a way to go. It has
functions to open an existing project file and explore, modify,
run, and save it from a Python script. There is a simple example
for the `tui` use at
https://github.com/diffpy/diffpy.pdfgui/blob/master/doc/tutorial/tui_mno_bond_lengths.py

For a more advanced use you'd need to check the API functions in
the diffpy.pdfgui.control module; in particular those provided
by the Fitting, FitDataSet and FitStructure classes.

Hope this helps,

Pavol

--
Dr. Pavol Juhas
Computational Science Initiative
Brookhaven National Laboratory
P.O. Box 5000
Upton, NY 11973-5000

Fred marlton

unread,
Nov 1, 2017, 5:43:28 PM11/1/17
to diffpy-users
Hi Pavol,

Thanks for the response.

Ok cool, so I should be able to get the exact same answer between the two. So there must be something wrong in my script.....

Awesome, thanks for that, I'll have to give the 'tui' a go sometime. 

Fred
Reply all
Reply to author
Forward
0 new messages