Tutorial on nano crystal diffraction

[n01r]

Trying the tutorial on nano-crystal diffraction and reporting issues.

• importing CrystFELPhotonDiffractor, CrystFELPhotonDiffractorParameters, PhotonBeamParameters is not sufficient to create the beam variable
• the units are missing due to the upgrade to using the pint python package
• [x] fix the imports
• [x] fix the beam object with units
• the link to the geometry file is broken
• [x] fix the geometry file link
• creating the parameters object throws a warning even though the geometry file exists #159
• [x] fix the tutorial: constructor of parameters object takes detector_geometry instead of geometry, thanks @belfhi

[n01r]

When trying to run the last command diffr_diagnostics.py -P -p1 -z True xstal_diffr.h5 I am encountering the following error:

Traceback (most recent call last):
  File "/data/netapp/s2e/simex/bin/diffr_diagnostics.py", line 25, in <module>
    from SimEx.Analysis.DiffractionAnalysis import DiffractionAnalysis, plt
  File "/data/netapp/s2e/simex/lib/python3.4/SimEx/__init__.py", line 33, in <module>
    from .Analysis.DiffractionAnalysis import DiffractionAnalysis
  File "/data/netapp/s2e/simex/lib/python3.4/SimEx/Analysis/DiffractionAnalysis.py", line 23, in <module>
    from SimEx.Analysis.AbstractAnalysis import AbstractAnalysis, plt, mpl
  File "/data/netapp/s2e/simex/lib/python3.4/SimEx/Analysis/AbstractAnalysis.py", line 32
    class AbstractAnalysis(object, metaclass=ABCMeta):
                                            ^
SyntaxError: invalid syntax

That is likely due to the #!/usr/bin/env python2.7 shebang in the beginning and the python3 branch being installed on Maxwell.

I replaced it by #!/usr/bin/env python in the local copy of diffr_diagnostics.py because the simex module loads python 3.4 anyway. ../simex_platform/Sources/python/ScriptCollection/DataAnalysis/scattering/diffr_diagnostics.py -P -p1 -z True xstal_diffr.h5

/usr/lib64/python3.4/importlib/_bootstrap.py:321: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.
  return f(*args, **kwds)
initializing ocelot...

WARNING: esther_execute could not be imported. This is most probably due to Esther not being installed or not found. Expect RunTimeErrors when attempting to run the EstherPhotonMatterInteractor.backengine().
WARNING: Giving an operation with a single pattern has no effect.

The resulting plot looks like this: That's not what it's supposed to look like, according to the tutorial :sweat_smile:

cc'ing @CFGrote

[rshopa]

Same here - looks like detector output contains low statistics, as @n01r noticed. Only 23 non-zero pixels of the same intensity.

[rshopa]

UPD: no improvement on diffraction patterns (i.e. few non-zero pixels only) if using files from SPI tutorial as input.

Unfortunately, I've found no way to "feed" CrystFEL (CrystFELPhotonDiffractorParameters() and CrystFELPhotonDiffractor()) with pmi_output - data with different structure of .h5 file. @CFGrote , is there any way to overcome this problem? I mean - compare the images obtained using SingFEL and CrystFEL after PMI for, say, the same sample 5udc.pdb?