punzo / slicerastro Goto Github PK

Three options:

1. Change the NaN values to zeroes at the loading of the datacube
2. Change the NaN values to zeroes before applying any operations (such as smoothing)
3. Masking all the values during the operations (expensive)
4. Provide a module and leave the conversion to the user

if NV is zero, barolo estimates it and it will use the new value in the fit. However, the value of the NV variable in the Param class is not updated. Therefore, SlicerAstro will show in the interface always zero, even if NV has been calculated internally.

The classes vtkOpenGLAstroShaderComputation and vtkOpenGLAstroTextureImage will not work with OpenGL2 (VTK7) (vtkgl classes not available anymore). Working on new version for the GPU filters based on Kitware/VTK@c3a6dfa

When the 3D rendering is zoomed inside the extent of the datacube (pink frame) the cloud lasso selection tool does not work.

In the astroVolume GUI

check if any updated is needed

Before calculating the WCS, the offset has to converted properly from RAS to IJK

See https://github.com/SBU-BMI/SlicerOpenCV/ for example

https://www.slicer.org/wiki/Documentation/Labs/Qt5-and-VTK8
some classes will need to be updated

See

reported by @Fil8

reported Anja Schroeder (PHISCC 2017)

https://github.com/Punzo/SlicerAstro/blob/master/AstroVolume/MRML/vtkMRMLAstroTwoDAxesDisplayableManager.cxx#L315

crush when loading a scene

different box color along the velocity axis (reported by Marc).

The copy method works, most likley the bug is in the SubjectHierarch classes

https://www.slicer.org/wiki/Documentation/Nightly/Developers/Plots

It is not clear if it is due to a patch to the wcs 5.11 to 5.14.

Fixing the header, removing all the fourth keywords may fix it.

fits_delete_key
https://heasarc.gsfc.nasa.gov/docs/software/fitsio/c/c_user/node39.html

and modify NAXIS:
https://heasarc.gsfc.nasa.gov/docs/software/fitsio/cexamples/modhead.c

This https://github.com/Punzo/Bbarolo needs this update
editeodoro/Bbarolo@55b0dd5

(Then let's merge to https://github.com/editeodoro/Bbarolo/)

Qt5 version of the Slicer and SlicerAstro binaries will be soon released. Therefore, this won't be fixed.

If in a scene there is an AstroVolumeNode and the import scene has an AstroVolumeNode as well (with same IDs), it results in a crash. The crash seems connected with AstroVolumeDisplayNode. It appears to be a reference issue.

vtkMRMLAstroVolumeNode::UpdateNoiseAttributes and UpdateRangeAttributes do not mask for NAN values.

Slicer/Slicer#540

see
Slicer/Slicer@b2bfa60

cmakeLists.txt should be updated to use the macosx packing issue fixed in Slicer r25182

loading initial cache file /.../SlicerAstro-build/inner-prefix/tmp/inner-cache-Release.cmake
-- Setting OSX_ARCHITECTURES to 'x86_64' as none was specified.
-- Setting OSX_SYSROOT to latest '/Developer/SDKs/MacOSX10.6.sdk' as none was specified.
-- Configuring incomplete, errors occurred!

CMAKE_OSX_DEPLOYMENT_TARGET must be 10.6 or greater.
Call Stack (most recent call first):
/Users/kitware/Dashboards/Nightly/Slicer-0-build/Slicer-build/SlicerConfig.cmake:106 (include)
CMakeLists.txt:28 (find_package)

vtkDebugLeaks has detected LEAKS!
Class "vtkXOpenGLRenderWindow" has 1 instance still around.
Class "vtkCellData" has 1 instance still around.
Class "vtkOpenGLPainterDeviceAdapter" has 1 instance still around.
Class "vtkOpenGLAstroShaderComputation" has 1 instance still around.
Class "vtkPointData" has 1 instance still around.
Class "vtkInformation" has 1 instance still around.
Class "vtkInformationIntegerPointerValue" has 1 instance still around.
Class "vtkOpenGLAstroTextureImage" has 2 instances still around.
Class "vtkOpenGLExtensionManager" has 1 instance still around.
Class "vtkIdList" has 1 instance still around.
Class "vtkInformationIntegerValue" has 1 instance still around.
Class "vtkRendererCollection" has 1 instance still around.
Class "vtkImageData" has 1 instance still around.
Class "vtkFieldData" has 1 instance still around.
Class "vtkCommand or subclass" has 1 instance still around.

Part of the code regarding the ruler has shifted from Python to C++. This means that SlicerAstro WCS axis has to be refactorized in C++.

1. // Setup Home Module
   cout<app->settings()->value("Modules/HomeModule").toString().toStdString()<<endl;
   d->app->settings()->setValue("Modules/HomeModule", "AstroWelcome");
   cout<app->settings()->value("Modules/HomeModule").toString().toStdString()<<endl;

2. // Setup modules to ignore
   qSlicerCoreApplication * coreApp = qSlicerCoreApplication::application();
   qSlicerAbstractModuleFactoryManager* factoryManager = coreApp->moduleManager()->factoryManager();
   QStringList modulesNames;
   modulesNames << "DICOM";
   factoryManager->setModulesToIgnore(modulesNames);

tried this in qSlicerAstroVolumeModule::setup()

1. it was changing the value to the AstroWelcome, but then the Home Module loaded was still the SlicerWelcome
2. using this Slicer get completely stucked.

another way can be setting up in slicerExtension a script after the installation of the extension
http://na-mic.org/Mantis/view.php?id=2920#c13958

crush if one of the FWHMx,y,z or Nx,y,z are set to zero (on GPU).

Adding a button to set a second 3D view (camera perpendicular) that shows the depth coordinate of the selection.

Hi @jcfr, would be possible to update the linux factory machine gcc from 4.6 to 4.7 to support OpenMP call reduction max and min? gcc 4.7 supports OpenMP 3.1 which is the one I need.

Or may you keep in mind this when setting up the new infrastructure to compile the Qt5/VTK9 version?

build Error:
#pragma omp parallel for schedule(static) reduction(max : max_val), reduction(min : min_val)
http://slicer.cdash.org/viewBuildError.php?buildid=1186462