xrayutilities-users — User mailinglist for xrayutilities

[Xrayutilities-users] xrayutilities 1.3.1 released

[Dominik K. <dom...@gm...>]

xrayutilities 1.3.1 was released (2016-05-19)

In this release xrayutilities gained the ability to simulate reflectivity
and diffraction curves for thin film samples based on kinematical and
dynamical diffraction models. This feature was already introduced in a
previous message to this list [1].

Further noteworthy changes are:
*) the new Crystal and Amorphous material classes replace the former
Material class, which should not be used anymore, and serves only as a
super class of the two mentioned types of materials. The Amorphous
materials are especially useful for the definition of materials for
reflectivity calculations, which do not need the information about the
crystal structure.
*) A lot of code cleanup was performed using clonedigger to identify
duplicated code
*) the peak_fit functions ability to guess the initial parameters was
heavily improved.
*) lots of small bug fixes

Please have a look to the new optional dependencies in the README if you
plan to use all the new features. The unittest code coverage was increased
to 61%.
See the CHANGES.txt file or the GIT log for a more complete list of changes.

As usual every user is advised to upgrade. In order to obtain the new
version go to sourceforge [2] or use PyPI [3]. Windows binaries are found
on PyPI for 64-bit Python in versions 2.7 and 3.5. In case a binary version
yo would need is missing or you find problems with this new version please
report them!

[1] https://sourceforge.net/p/xrayutilities/mailman/message/35024338/
[2] https://sourceforge.net/projects/xrayutilities
[3] https://pypi.python.org/pypi/xrayutilities

Re: [Xrayutilities-users] New X-ray reflectivity and diffraction simulation-code

[Dominik K. <dom...@gm...>]

Dear xrayutilities users,

I have updated the release candidate [1] which now includes much faster
calculation of the diffraction signal. Especially the dynamical diffraction
model increased its speed more than 25times due to significant optimization
work. Kinematical diffraction models also decreased their calculation time
by roughly a factor of 2.

after one more round of testing the final release will take place soon.

regards
Dominik

[1] release candidate 1.3rc2 release is available at
https://sourceforge.net/projects/xrayutilities/files/

On Tue, Apr 19, 2016 at 12:48 PM, Dominik Kriegner <
dom...@gm...> wrote:
>
> Dear xrayutilities users,
>
> I want to announce a new feature of xrayutilities which will be included
in the upcoming release 1.3.
>
> Since xrayutilties already for a long time includes a database of
materials and is able to calculate their optical properties it seemed like
an obvious addition to add some simulation abilities which need those
material parameters as input.
>
> Within the current development version [1,2] such support for x-ray
reflectivity and x-ray diffraction calculation for thin films was added.
Several different models for diffraction are available and documented [3].
>
> Clearly more testing of the code is needed to ensure its correct function
in all use cases. I encourage you to try the new abilities and let me know
any feedback/problems you might have. In particular interesting for me is
if the documentation and new example-scripts are sufficient for you to get
started with your simulations!
>
> regards
> Dominik
>
> [1] https://sourceforge.net/p/xrayutilities/code/ci/master/tree/
> [2] a release candidate for the 1.3 release is available at
https://sourceforge.net/projects/xrayutilities/files/
> [3] http://xrayutilities.sourceforge.net/simulations.html

[Xrayutilities-users] New X-ray reflectivity and diffraction simulation-code

[Dominik K. <dom...@gm...>]

Dear xrayutilities users,

I want to announce a new feature of xrayutilities which will be included in
the upcoming release 1.3.

Since xrayutilties already for a long time includes a database of materials
and is able to calculate their optical properties it seemed like an obvious
addition to add some simulation abilities which need those material
parameters as input.

Within the current development version [1,2] such support for x-ray
reflectivity and x-ray diffraction calculation for thin films was added.
Several different models for diffraction are available and documented [3].

Clearly more testing of the code is needed to ensure its correct function
in all use cases. I encourage you to try the new abilities and let me know
any feedback/problems you might have. In particular interesting for me is
if the documentation and new example-scripts are sufficient for you to get
started with your simulations!

regards
Dominik

[1] https://sourceforge.net/p/xrayutilities/code/ci/master/tree/
[2] a release candidate for the 1.3 release is available at
https://sourceforge.net/projects/xrayutilities/files/
[3] http://xrayutilities.sourceforge.net/simulations.html

[Xrayutilities-users] xrayutilities 1.2.1 released

[Dominik K. <dom...@gm...>]

xrayutilities 1.2.1 was released (2018-01-08)

The most significant addition in this release is clearly the availability
of fuzzy gridders for 1, 2, and 3 dimensional data which was already
introduced separately before [1].
In addition several file parser (FIO, cif, TIFF) were improved and support
for Rigaku (RAS) files was added. The unittest code coverage was increased
to 57%.

See the CHANGES.txt file or the GIT log for a more complete list of changes.

As usual every user is advised to upgrade. In order to obtain the new
version go to sourceforge [2] or use PyPI [3]. Windows binaries are found
on PyPI for 32 and 64-bits and Python version 2.7 and 3.4. In case you find
problems with this new version please report them!

[1] http://sourceforge.net/p/xrayutilities/mailman/message/34621608/
[2] https://sourceforge.net/projects/xrayutilities
[3] https://pypi.python.org/pypi/xrayutilities

[Xrayutilities-users] fuzzy gridding and reciprocal space map visualization

[Dominik K. <dom...@gm...>]

Dear xrayutilities-users,

recently I implemented a new way of visualizing reciprocal space maps which
might be useful for several of you so I decided to post a small explanation
here.

For the visualization of reciprocal space maps the irregularly spaced
experimental data are often "gridded" onto a regular grid. Although this 
sometimes causes problems [1] its often simply necessary.
Especially when any loss of resolution is to be avoided the grid density has to
be chosen at limiting values which often cause artifacts in the data. Such
artifacts are for example seen in the left most panel of the attached figure.

To make this process more smooth I implemented something I call
"fuzzy"-gridding.  For the 1D case thats described in the appendix of an
article about powder diffraction we published earlier this year [2]. The basic
idea is to mimic better the experimental conditions in which every data point
is actually not a mathematical point but has to be considered to have some
extension due to the limited experimental resolution. Considering this
resolution in the gridding process by distributing data points fractionally on
multiple grid points on can avoid the girdding artifacts very easily (see
middle panel of the attached figure).

For comparison the figure also shows the visualization with a matplotlib
function 'pcolormesh' which in certain cases allows to avoid gridding
all-together. see its documentation for details [3].

The code and data to generate the attached image can be found in the example
directory of xrayutilities (see GIT master [4]). Fuzzy gridding functions are
now available for 1D, 2D and 3D gridders and will be inlcuded in the next
release of xrayutilities.

Any feedback is welcome!

cheers
dominik

[1] https://sourceforge.net/p/xrayutilities/mailman/message/32836996/
[2] http://scripts.iucr.org/cgi-bin/paper?S1600576715003465
[3] http://matplotlib.org/api/pyplot_api.html#matplotlib.pyplot.pcolormesh
[4] https://sourceforge.net/p/xrayutilities/code/ci/master/tree/examples/xrayutilities_fuzzygridding.py

[Xrayutilities-users] database update for xrayutilities 1.2.0

[Dominik K. <dom...@gm...>]

Dear all,

recently my attention was drawn to the fact that in the InterTables
database files the different oxidation states/radicals of the atoms are
listed separately. All available data from this database are now available
in xrayutilities (these changes are included in xrayutilities from version
1.2.0 on)

For example Mn2+ is found as

  import xrayutilties as xu
  xu.materials.elements.Mn2p

Radicals are denoted by a 'dot' in the name, e.g. the radical O2-. is found
as

  xu.materials.elements.O2mdot

At the same time I cleaned up a bit the code dealing with this database and
could significantly improve the speed of the database access.

In course of this I switched from using pytables to h5py for the HDF5 file
access. This means you might need to install this package. It is, however,
likely that you have it already. pytables will be no longer required!

cheers
dominik

ps: here some of the incredible differences of time measurements before and
after the recent changes:

before the update
------------------------
%time xu.materials.elements.Mn.f0(linspace(0,10,10000))
CPU times: user 704 ms, sys: 0 ns, total: 704 ms
Wall time: 705 ms

%time xu.materials.elements.Mn.f1(linspace(1000,10000,100000))
CPU times: user 21.9 s, sys: 4 ms, total: 21.9 s
Wall time: 21.9 s

after the update
---------------------
%time xu.materials.elements.Mn.f0(linspace(0,10,10000))
CPU times: user 4 ms, sys: 0 ns, total: 4 ms
Wall time: 5.74 ms
For this case the speedup is a factor of ~100

%time xu.materials.elements.Mn.f1(linspace(1000,10000,100000))
CPU times: user 4 ms, sys: 0 ns, total: 4 ms
Wall time: 4.39 ms
For this maybe slightly academic case this is a speedup of 5000!!!

[Xrayutilities-users] xrayutilities 1.2.0 released

[Dominik K. <dom...@gm...>]

xrayutilities 1.2.0 was released (2015-08-03)

In this version a lot of things were changed below the hood. Especially
there is a new dependency (h5py), which replaces pytables for all HDF5 file
accesses. The database with the scattering factors for the chemical
elements was updated and now works significantly faster.
Make sure you have h5py installed, it is, however, anyhow included in
popular Python distributions (e.g. WinPython and Python(x,y), or use the
package python-h5py on Ubuntu/Debian).

The detailed changes since the last release include:
* improved code quality due to increased use of unittests (code coverage:
49%) -> if you want to run the test code (not required) you should download
the test data files from [1]
* Radicon file support was removed (no test scripts available!)
* replace pytables by faster and more convenient h5py
* bug fixes/improvements in the file parsers (discovered by the unit tests!)
* added different oxidation states of the chemical elements to the database
(separate announcement will follow on the mailing list)
* memory usage optimization in FastScanCCD

See the CHANGES.txt file or the GIT log for a more complete list of
changes.

As usual every user is advised to upgrade, however, this time take care
that you have the new dependency (h5py) installed! In order to obtain the
new version go to sourceforge [2] or use PyPI [3]. Windows binaries are
found on PyPI for 32 and 64-bits and Python version 2.7 and 3.4. In case
you find problems with this new version please report them!

[1] https://sourceforge.net/projects/xrayutilities/files/?source=navbar
[2] https://sourceforge.net/projects/xrayutilities
[3] https://pypi.python.org/pypi/xrayutilities

[Xrayutilities-users] xrayutilities 1.1.2 released

[Dominik K. <dom...@gm...>]

xrayutilities 1.1.2 was released (2015-04-22)

The changes since the last release include:

* a new way of reading spec scan data, which works without using HDF5 files
  and reads only selected scans and can therefore be more efficient.
  The new function is named xu.io.getspec_scan
* huge speed improvements when reading EDF files with multiple images per
file
* A lot of improvements when working with FastScan data (e.g. acquired
using the
  kmap mode at ID01). See the example script
xrayutilities_kmap_example_ESRF.py
* last but not least: a shiny new mobile friendly webpage [1] generate with
the
  sphinx-rtd-theme [2]

See the CHANGES.txt file or the GIT log for a more complete list of changes.

As usual every user is advised to upgrade! In order to obtain the new
version go to sourceforge [2] or use PyPI [3]. Windows binaries are found
on PyPI for 32 and 64-bits and Python version 2.7 and 3.3 & 3.4. In case
you find problems with this new version please report them!

[1] http://xrayutilities.sourceforge.net/
[2] https://github.com/snide/sphinx_rtd_theme
[3] https://sourceforge.net/projects/xrayutilities
[4] https://pypi.python.org/pypi/xrayutilities

[Xrayutilities-users] xrayutiliies 1.1.1 released

[Dominik K. <dom...@gm...>]

xrayutilities 1.1.1 was released (2015-01-22)

Some new features include

*) new FuzzyGridder1D allowing to assign every data-point a certain width
which will distribute them over multiple output bins according to the
overlap.
*) the reciprocal space conversion (for point, linear, area detectors) can
now handle translation-stages in the detector arm
*) the reciprocal space conversion (for point, linear, area detectors) can
also treat now energy scans, where instead (or in addition) to angles the
energy of the x-ray photons is altered.
*) new PseudoVoigt1dasym function implementing an asymmetric PseudoVoigt
peak

This release contains also a lot of improvements which are not visible to
the user on first sight:

*) improvements in the file parsers, which now keep the files closed when
not parsing them, which was a weak point so far.
*) a memory leak in the XRDML file parser was closed by changing the backend
*) better automatic determination of the intial parameters in the peak_fit
function

See the CHANGES.txt file or the GIT log for a more complete list of changes.

As usual every user is advised to upgrade! In order to obtain the new
version go to sourceforge [1] or use PyPI [2]. Windows binaries are found
on PyPI for 32 and 64-bits and Python version 2.7 and 3.3. In case you find
problems with this new version please report them!

[1] https://sourceforge.net/projects/xrayutilities
[2] https://pypi.python.org/pypi/xrayutilities

[Xrayutilities-users] xrayutilities 1.1.0 released

[Dominik K. <dom...@gm...>]

xrayutilities 1.1.0 was released (2014-10-09)

This release includes several new features already highlighted before.
Since the release of the last beta version as few bugs where fixed,
the documentation was updated and lots of code quality improvements
were added. Every user is advised to upgrade!

In order to obtain the new version go to sourceforge [1] or use PyPI
[2]. Windows binaries are found on PyPI for 32 and 64-bits and Python
version 2.7.

In case you find that this version does not work for you or there are
any issues feel free to report them to this mailing list!

[1] https://sourceforge.net/projects/xrayutilities
[2] https://pypi.python.org/pypi/xrayutilities

[Xrayutilities-users] xrayutilities 1.1.0-beta released

[Dominik K. <dom...@gm...>]

Today (2014-09-16) a pre-release of the upcoming version 1.1.0 of
xrayutilities was released.

This function includes a lot of improvements and fixes and every user is
advised to update to this version. The release is called 'beta' in order to
wait a few weeks to find remaining bugs before the release of the final
1.1.0, but I do not expect major changes.

A detailed summary of the changes since the last release is found in the
Changelog. Following are just the most important changes:

* a new parser for FastScan data from spec-mesh scans and FastScan/kmap
data from ESRF was added (developed together with Raphael G.)

* a lot of Documentation updated triggered by user feedback; thanks for
that!

* improved python 3.X compatibility and more test code to automatically
reveal errors in core parts of the package

In order to obtain the new version go to sourceforge [1] or use PyPI [2].
Windows binaries are found on PyPI for 32 and 64-bit and Python2.7 and 3.3,
respectively.

[1] https://sourceforge.net/projects/xrayutilities
[2] https://pypi.python.org/pypi/xrayutilities

[Xrayutilities-users] update to version 1.0.5 -> changes in Gridder classes

[Dominik K. <dom...@gm...>]

Dear XrayUtilities-Users

Lately there were some changes in XrayUtilities which are part of version
1.0.5 (released today: February 9th 2014)  which might make small changes
to your scripts necessary.

In particular the changes affect you if you are using the *Gridder* classes
of xrayutilities

Let me briefly outline how the Gridder now works. The following basically
applies to 1D,2D, and 3D gridder in the same way. I will use the 2D gridder
as an example.

The most easiest use (what most user might need) is:

import xrayutilities as xu # import python package
g = xu.Gridder2D(100,101) # initialize the Gridder object
#====== load some data here =====
g(x,y,data) # call the gridder with the data
griddata = g.data # the data object of the Gridder contains the gridded
data.

Note: previously you could use the g.gdata data object which was always an
internal buffer and should not be used anymore!

A more complicated example showing also sequential gridding is shown below.
You need sequential gridding when you can not load all data at the same
time. in such cases you need to specify the data range before the first
call to the gridder.

import xrayutilities as xu # import python package
g = xu.Gridder2D(100,101) # initialize the Gridder object
g.KeepData(True)
g.dataRange((1,2),(3,4))  # (xgrd_min,xgrd_max),(ygrd_min,ygrd_max)
#====== load some data here =====
g(x,y,data) # call the gridder with the data
griddata = g.data # the data object of the Gridder contains the so far
gridded data.

#====== load some more data here =====
g(x,y,data) # call the gridder with the new data
griddata = g.data # the data object of the Gridder contains the combined
gridded data.

In the new version there are also be some more improvements and small bug
fixes and I recommend everyone to update to this version. In particular
interesting is that now you can now read basically all data files directly
from zipped files (gzip and bzip2).

kind regards

Dominik

[Xrayutilities-users] xrayutilities v1.0

[Dominik K. <dom...@gm...>]

Hi all,

Today I uploaded version 1.0 of xrayutilities to sourceforge and to the 
python package index. 
As announced earlier this release includes major changes and I recommend 
that every user should update to this release. It also includes several 
fixes to make xrayutilities run on more MS windows machines. 

The downloads can be found at:
https://sourceforge.net/projects/xrayutilities/ 
(source tarball for all platforms)
or
http://pypi.python.org/pypi/xrayutilities/  
(executable installer for windows + source tarball)

A description of xrayutilities can be found at the updated webpage 

http://xrayutilities.sourceforge.net/
or in a recent open access article published in Journal of Applied
Crystallography, which can be found at
http://dx.doi.org/10.1107/S0021889813017214

In case you encounter any problems please let me know.

regards
Dominik

-- 
mail: dom...@gm...
gpg: FA4B630F

[Xrayutilities-users] conversion of c-library to c-extension

[Dominik K. <dom...@gm...>]

Dear xrayutilities users

During the last month a lot of things changed in xrayutilities.
We (Eugen and me) converted the c-library to a Python c-extension.
This sounds like a minor change but it makes xrayutilities a proper
python package, which can be distributed and installed more easily.

The most basic installation is now done using

  python setup.py install

or

  python  setup.py install --prefix=<installpath>

To use the new version 0.99 download it from sourceforge or pull it
from the GIT repository.
For detailed installation instructions look into the readme file.
Converting existing installations can be done by just installing the
new package.

Based on the current code we plan to release xrayutilities-1.0 soon.
In case you find any bugs or have done any changes which you think are
worth inclusion in this release please let us know within the upcoming
weeks.

Along with the change of the c-code we updated the Python code and
added some new feature regarding parallelization and detector
parameter determination and some basic unittests for core
functionality to ensure that they work and do not get broken by code
regressions.
The detector parameter determination for area detectors can now be
done with data from scans in the primary beam and at Bragg peaks of a
known material measured in coplanar diffraction. This significantly
increases the sensitivity to the "outer angle offset", which was
previously quite pure for larger detector distances.
The number of threads used by the parallel section of the code can now
easily be changed in the xrayutilities configuration.
More details about the configuration and other features can be found
in the updated documentation (http://xrayutilties.sf.net)

regards

Dominik

[Xrayutilities-users] xrutils -> xrayutilities

[Dominik K. <dom...@gm...>]

Dear Users,

approximately two weeks ago the python package of xrayutilities was
renamed from

xrutils

to

xrayutilities

to make everything more consistent,

This means if you recently updated or use the git code you have to
rename all your imports in all your scripts. Since this can be tedious I
attach a script which can (on Linux/Unix) do this automatically.

Please use it carefully since it will change all the python files it
finds! I can not give any warranty that this works, so use it on your
own risk!

One my machines it works flawlessly and did all the necessary changes.

regards

Dominik

[Xrayutilities-users] new webpage with documentation

[Dominik K. <dom...@gm...>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hello all,

Just to let you know since a few days there is a new webpage with the
documentation of xrayutilities. There was already plenty of
documentation available but it was not so easy to access. I hope that
with this webpage it will get easier to get started.

you can find it at:

xrayutilities.sourceforge.net
xrayutilities.sf.net

Also plenty of work was put in xrayutilies during the last few weeks.
Most of it is not directly visible to user, do however consider
updating to the newest version to get the latest changes.

regards

Dominik
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.19 (GNU/Linux)
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=e1uS
-----END PGP SIGNATURE-----

[Xrayutilities-users] code-change: return values for angle calculation

[Dominik K. <dom...@gm...>]

Hi all,

recently the return values of the Q2Ang functions of the Experimental
classes were changed with the attempt to make them more complete and
more uniform. This makes code changes in your scripts necessary. See the
help or example scripts or below for details.

As an example the HXRD class now must be used like:

import xrutils as xu
Si = xu.materials.Si
exp = xu.HXRD(Si.Q(1,-1,0),Si.Q(0,0,1),en = 9000)

exp.Q2Ang(Si.Q(0,0,4))
#returns: array([ 30.48461232,   0.  ,   0.  ,  60.96922465])
exp.Q2Ang(Si.Q(2,2,4),geometry='hi_lo')
#returns: array([ 73.67709683,   0.  ,  90.  ,  76.8254143 ])
exp.Q2Ang(Si.Q(2,2,4),geometry='lo_hi')
#returns: array([  3.14831747,   0.  , -90.  ,  76.8254143 ])

it returns the angles om,chi,phi,tt (sample rocking angle, sample tilt
perpendicular to scattering plane, sample azimuth, scattering angle two
theta)

Furthermore there is the option to return the angles needed for
non-coplanar scattering geometry using the geometry 'realTilt' or the
NonCOP experimental class.

exp.Q2Ang(Si.Q(2,2,4),geometry='realTilt')
#returns: array([ 38.41270715,  35.26438968,   0.  ,  76.8254143 ])

expn = xu.NonCOP(Si.Q(1,-1,0),Si.Q(0,0,1),en = 9000)
expn.Q2Ang(Si.Q(2,2,4))
#returns: array([ 38.41270715,  35.26438968,   0.  ,  76.8254143 ])

There are also specialized classes for grazing incidence diffraction
measurements. See help of xu.GID and xu.GID_ID10B for details.

regards

dominik

[Xrayutilities-users] Welcome to the list!

[Dominik K. <dom...@gm...>]

Hello List!

(In fact empty list at the moment!)
This list should serve as common place for announcements regarding
important changes of xrayutilities. 

This should include
* announcements of new versions with significant changes
  along with a call for user to update to the new version
* announcements of important bug-encounters and fixes
* general discussion regarding xrayutilities

The list will most probably be low-traffic!

kind regards

Dominik