summaryrefslogtreecommitdiff
path: root/changelog.txt
diff options
context:
space:
mode:
authorPicca Frédéric-Emmanuel <picca@debian.org>2012-05-02 22:04:26 +0200
committerPicca Frédéric-Emmanuel <picca@debian.org>2012-05-02 22:04:26 +0200
commit800d5117dc26c5a2ad2fc8bd966fe1f631076e88 (patch)
tree61f9f72875e4db87c76a849b7ed57851219a24f9 /changelog.txt
parent5b5ffc94d548209f3d15be3c457417c39392dd96 (diff)
Imported Upstream version 4.6.0~rc1
Diffstat (limited to 'changelog.txt')
-rw-r--r--changelog.txt604
1 files changed, 604 insertions, 0 deletions
diff --git a/changelog.txt b/changelog.txt
new file mode 100644
index 0000000..1bcf814
--- /dev/null
+++ b/changelog.txt
@@ -0,0 +1,604 @@
+VERSION 4.5.0
+
+Support h5py version 2.x
+
+Support OMNIC 8.x .map file format
+
+Support TIFF format. Uncompressed and packbits.
+
+Support a couple of ASCII based file formats of beamlines of Diamond and APS.
+
+Preliminary Python 3.2 support.
+
+Support principal components analysis of dynamically loaded stacks.
+
+Improved HDF5 file format handling.
+
+Possibility to visualize variable width line profiles on images.
+
+Add split Gauss, split Lorentz and split Pseudo-Voigt as fitting functions.
+
+Do not automatically sort the list of files to be treated in batch mode. The
+program will respect the order provided by the user.
+
+Allow visualized 3D objects to share the same colormap.
+
+Lower the energy threshold in order to consider Boron K X rays.
+
+Correct Arsenic density.
+
+Allow to customize the saving of Scan window curves using matplotlib.
+
+Allow multiple processes in MacOS X when using a non frozen PyMca version.
+
+VERSION 4.4.1
+
+Allow detailed customization of 1D output.
+
+Add simple normalization plugins to the Scan window.
+
+Implement a plugin system for the ROI Imaging tool.
+
+Implement a generic batch fitting of stacks as a stack plugin. The functions are totally specified by the user.
+
+Implement simple background removal tools as stack plugins.
+
+Implement an alternative ROI window as a stack plugin.
+
+Support convertion of big EDF image stacks to HDF5.
+
+Visualization of large image stacks via dynamic loading and use of HDF5.
+
+Support MDP based PCA and ICA on user selected regions.
+
+Read calibration from OMNIC 7.x .map files.
+
+Support Bruker Opus maps exported in DPT format.
+
+Improve the reading speed of TwinMic .dta files.
+
+Add simple MarCCD support.
+
+Add very simple support for some flavours of Pilatus CBF.
+
+Show file header information.
+
+Properly handle Amptek MCA calibration when more than two points or ROIs have been used.
+
+Allow Savitzky-Golay (SG) filtering of stacks.
+
+Correct problem affecting odd order SG derivatives.
+
+Very primitive HDF5 2D Visualization.
+
+Possibility to visualize HDF5 3D datasets as series of images.
+
+Implement a context menu on the main window HDF5 browser.
+
+Table visualization of non-numerical HDF5 datasets.
+
+Support segmented HDF5 files using default segmentation.
+
+llow the use of several processes when batch fitting a single HDF5 file.
+
+Allow reading of pure image formats in the RGB correlator.
+
+The fit configuration window was too high for Mac computers having exactly 800
+pixels vertical resolution.
+
+Workaround a windows problem when trying to select thousands of files in the
+batch file dialog by using Qt file dialogs.
+
+Minor visualization and data handling improvements.
+
+Prevent a crash when the excitation energy is below 1 keV.
+
+Add a script to build PyMca from source on Debian or Ubuntu systems.
+
+Make specfile LC_NUMERIC locale independent to solve Debian bug report 602471.
+
+Unify the sps library used by PyMca and the one used by Certified Scientific
+Software package SPEC. Use a BSD like license for the associated code.
+
+VERSION 4.4.0
+
+Extend the usability of the code down to 100 eV extending XCOM mass
+attenuation coefficients with EPDL97 data when energies are below 1 keV.
+
+Alternative background subtraction algorithm (SNIP) implemented.
+
+Support of HDF5 file format.
+
+Possibility to export ROI imaging tool data stack in HDF5.
+
+Workaround 32-bit limit when handling huge EDF maps in 32-bit machines.
+
+Better customization capabilities when saving images through matplotlib.
+
+Support Fit2D .chi files.
+
+Simple support of ADSC files wrapping them as EDF files.
+
+Possibility to save images as 32-bit floats for people using the ImageJ EDF
+plugin.
+
+Support combined PCA when using two data stacks.
+
+Support Independent Component Analysis via MDP if installed.
+
+Support non-negative Matrix Approximation Analysis using Uwe Schmitt modules
+(http://public.procoders.net/nnma/)
+
+Add multivariate analysis capabilities to the RGBCorrelator.
+
+Implement a mechanism to add plugins to 1D plots.
+
+Implement simple 3D visualization capabilities.
+
+Solve X-ray tube profile generation problems at tube voltages below 20 kV.
+
+Allow a greater number of energies defining the X-ray tube emission profile.
+
+Correct logarithmic colormap handling.
+
+Correct colormap problems in 64-bit platforms.
+
+Correct problem preventing data calibration in some 64-bit platforms.
+
+Minor bugs corrected and minor features added.
+
+VERSION 4.3.0
+
+Speed up fit configuration when using multiple energies.
+
+Import/export the multiple energies beam description as CSV files.
+
+Possibility to calculate transmission curves added to the tools menu.
+
+Offer the possibility to use an additional filter in the x-ray tube setup.
+
+Visualization of the absorption and the detector contribution to the efficiency.
+
+Prevent the use of trailing spaces in the definition of materials.
+
+Implement zoomed window panning (press CTRL key and desired arrow key simultaneously).
+
+Offer contour plot customization when saving images though matplotlib.
+
+Possibility to flip external images used for selection in the ROI imaging
+tool.
+
+Numpy 1.2 runtime deprecation warnings suppressed.
+
+Interpolated mass attenuation coefficients were between 0 and 3 %
+overestimated.
+
+Correct a bug appearing when trying to fit separate Ka and Kb lines of
+elements above Rb.
+
+VERSION 4.2.6
+
+Add very basic spx file format support.
+
+Improved SPE file format support.
+
+Improved semilogarithmic Y axis toggling.
+
+Correct a problem appearing when adding curves to the scan window in show points only mode.
+
+Offer the possibility to choose between Qt and native file dialogs through the
+PyMca command line argument --nativefiledialogs=1.
+
+Allow overwriting ROI definition files.
+
+Two column ASCII files were not read in the ROI Imaging.
+
+Implement the ROI Imaging --fileindex=1 command line option to deal with the case
+the EDF map was column oriented instead of row oriented.
+
+Add the possibility to transpose the resulting images of a batch fitting by
+passing the --fileindex=1 option to PyMcaPostBatch.
+
+Add threshold selection methods.
+
+Implement selections from external images.
+
+Correct a bug affecting saving of McaAdvancedFit data in CSV format.
+
+Replace corrupted Ru.mat attenuation data.
+
+Solve a problem with fit configuration window size affecting MacOS computers.
+
+Prevent the use of numbers as material names.
+
+Prevent an endless loop in case of typing a bad matrix element in the concentrations tab.
+
+Add the necessary scripts to build PyMca as a python library under windows.
+
+The calibration was not properly written when converting to EDF file format.
+
+VERSION 4.2.5
+
+- Batch fitting.
+ Save concentrations as images too.
+
+- Advanced Fit changes.
+ Previous linear fit change was missing in batch mode.
+ Correct a problem showing up when Kr L lines were fitted as a group.
+
+- ROI Imaging.
+ Show the calculated principal components when using PCA.
+ Useless stack background subtraction was not working.
+
+VERSION 4.2.4
+- Add ESRF Xia Correct to the Tools menu.
+
+- ROI Imaging changes.
+ Do not reset ROI image selection on ROI change.
+ Add data saving capabilities to the ROI image window.
+ Implement basic principal component analysis mapping.
+ Support AIFIRA and SupaVisio maps.
+
+- Advanced Fit changes.
+ Improved transmission geometry support.
+ User can specify the scattering angle in the fit configuration.
+ Support "funny type" attenuators.
+ Make better use of the available space in the McaAdvancedFit window.
+ Linear fit: fix peaks to 0 area if they are outside the fitting region.
+
+- Simplify access to binding energies by putting them in a separate ASCII file.
+
+- Add anchors and smoothing information to the HTML reports.
+
+- Correct a bug on the reconstruction of the chisq image when using the
+multiple processes batch option.
+
+- Correct a bug preventing the generation of HTML reports in batch mode.
+
+- Correct a specfile compilation issue with some gcc versions.
+
+- Suppress shared memory compilation warnings.
+
+VERSION 4.2.3
+- Possibility to use pseudo-Voigt as advanced fit function.
+
+- Possibility to save as data and to remove selected images from
+RGBCorrelator.
+
+- Support Lucia beamline maps.
+
+- Support Diamond I18 beamline maps.
+
+- Correct ROI fitting option bug.
+
+- Some other minor bugs corrected and some other minor features added.
+
+VERSION 4.2.2
+
+- Input/Output: Support CSV format.
+ This should simplify data exchange with spreadsheets.
+
+- Input/Output: Support 64-bit handling of EDF files.
+
+- ScanWindow: Possibility to save all curves in a single specfile.
+
+- McaWindow: Spectra are saved with the active calibration instead of its own.
+ In other words: what you see is what you get.
+
+- Possibility to rename graph curves.
+
+- Possibility to save images thru matplotlib.
+ This allows user customization (axes, colorbar, colormap, contours, ...)
+
+- ROI imaging: Use file-mapped arrays for stacks of more than 2 Gbytes.
+ This feature is implemented for 64-bit linux systems.
+
+- McaAdvancedFit: The combination of linear fit and fixed pile-up was buggy.
+
+- Calibration: Correct a crash under python 2.5 using the Auto FWHM option.
+
+- RGB Correlator: ASCII images saved under windows can be read under linux.
+
+- Add peak statistics information to fit reports.
+
+- Some other minor bugs corrected and some other minor features added.
+
+VERSION 4.2.1
+
+- ROI image: Possibility to normalize spectra to number of pixels.
+
+- Speed up access to specfiles containing many scans.
+
+- Correct a bug that was making the calculation of net ROI images useless.
+
+- Correct ESRF specific data acquisition monitoring problems.
+
+VERSION 4.2.0
+
+- ROI Imaging Tool: Possibility to calculate net ROI images.
+
+- ROI Imaging Tool: Minimum width was a problem on dual screen displays.
+
+- ROI Imaging Tool: When using two stacks, the colormaps where coupled.
+
+- Main MCA Window old ROI markers were not erased under some conditions.
+
+VERSION 4.1.1
+
+- ROI Tool: Possibility to handle two stacks of same size simultaneously.
+ The interest is to combine fluorescence and diffraction maps.
+
+- Possibility to save ROI configuration directly from the table.
+
+- ROI Tool: Stack background removal implemented.
+ It may be of interest on diffraction maps. It is very slow and not worth
+ for fluorescence maps. The advanced fit is much better in this later case.
+
+- Linux: Add pymcaroitool to the binary package.
+ The ROI Tool can be started as an independent application and not only from
+ within PyMca.
+
+- Corrected: Simple fit combined with an energy calibration works.
+
+- Corrected: Scan monitoring. Last scan point was not shown.
+
+VERSION 4.1.0
+
+- Changed numerical library from Numeric to Numpy.
+
+- Added a tool to convert spectra to ESRF data format.
+
+- Improved shared memory access.
+
+- Possibility to calculate concentrations in mM.
+
+- Possibility to export the main MCA window in graphical format.
+
+- MCA window inactive legends after a fit problem corrected.
+
+- Corrected a problem in which HTML reports generated interactively
+after changing the concentration parameters without passing thru the
+fit configuration window could be different from those shown on the
+screen.
+
+VERSION 4.0.9
+
+- Concentration calculated on batch mode were wrong if the fundamental
+parameter method was employed and no report was generated. This bug was
+introduced in version 4.0.2
+
+- Batches were unable to start if the first file had a name with
+more characters than the last one.
+
+- Solved a random crash when showing the fit configuration dialog.
+
+- Program was unable to start under Python 2.5
+
+- Possibility to set anchors to the strip background. This makes the
+strip background the ideal choice when the spectra present background
+pile up artifacts.
+
+- Correct a memory leak appearing when the Savitsky-Golay smoothing was one.
+
+VERSION 4.0.8
+
+- Upgraded underlying sip and PyQt libraries.
+
+- Scan window operational.
+
+- Shared memory access operational.
+
+- Batches can profit from multiple processors (windows and linux).
+
+- Batch deadlocks solved by the library upgrade.
+
+VERSION 4.0.7
+
+- Improved ergonomy.
+ A lot of clicks saved ...
+
+- Improved handling of ini files.
+
+- A lot of work made on the scan window.
+ It is quite close to being fully operational.
+
+- Improved shared memory access.
+ Almost fully operational.
+
+- Solved funny crashes under windows.
+ The program was crashing under some particular situations
+ like using remote desktop, hiding the windows taskbar or
+ when the screensaver was activated.
+
+- Several minor bugs fixed.
+
+VERSION 4.0.6
+
+- Recover material editor functionality.
+
+- Solve a printing issue on some systems.
+ A black square was always present at the lower right corner.
+
+- Started to add functionalities to the scan window.
+
+VERSION 4.0.5
+
+- Improvements in the RGB Correlator/Post Batch tool:
+ Read new images.
+ Perform mathematical operations with the images.
+
+- Solve a file format identification problem.
+ Some EDF files where wrongly identified as specfile.
+
+- Correct a bug reading specfile formatted files.
+ In case of multiple scans, only the mca(s) of the first scan were read.
+
+- Clear all items from print preview working.
+
+
+VERSION 4.0.4
+
+- Add the possibility to save graphics in vectorial format (svg).
+
+- Scale print preview to have a width equal to 50% of the page width.
+
+- Solve a problem in specfile batches when dealing with the @CHANN entry.
+
+- Correct ROI image colormap bug in ROI imaging tool.
+
+- Add specfile support to the ROI imaging tool.
+
+- Correct random crashes on startup and on specfile batches.
+
+VERSION 4.0.3
+
+- Correct a bug in the ROI imaging tool.
+ If the first channel was different from 0, there was an offset in each ROI.
+
+- Windows: Correct a bug making the applications crash on startup.
+
+- Speed up fit configuration dialog. It was awfully slow since version 3.9.4 Qt4.
+
+- Linux: Matplotlib saving of graphics working in the binary release also under (K)Ubuntu.
+ (It was already the case in pymca4.0.2-linux-p1.tgz)
+
+VERSION 4.0.2
+
+- Allow the user to choose if the data have to be weighted during the fit.
+
+- Correct a bug only affecting linear batches: areas and concentrations were smaller by
+a factor 10000. The bug was introduced in version 4.0.0
+
+- Faster concentrations calculation.
+
+- Concentrations calculated on batches do not force the generation of the .fit files.
+
+- Linux: sps library made locale independent.
+
+VERSION 4.0.1
+
+- Plotting under windows should be faster.
+
+- Solved a small memory leak usign the strip background.
+
+- Advanced fit config button reported not to work on a gentoo 64-bit system when building PyMCA from source. Corrected.
+
+VERSION 4.0.0
+
+- Simplify batch preparation.
+ After a fit, you can load the non-linear parameters from last fit into the configuration just by pressing a button.
+
+- Huge speed up of pure imaging batches (specially linear fits).
+
+- Possibility to save the imaging tool generated images in ASCII and EDF formats besides pure graphical formats.
+
+- Display (X, Y, Z) coordenates in EDF viewers.
+
+- Mark selected peaks in fit window, not in the main McaWindow.
+
+- The windows installation allows to keep the old 3.9.4 version (ESRF)installed.
+ That windows version can be downloaded from the ESRF. It has less features but a faster interface.
+ Physics is not changed and files generated by the 3.9.4 version can be used with this version.
+
+- Repeated HTML reports were incorrectly visualized.
+
+- Correct a memory leak in the specfile library.
+
+- Speed up sequential mca readout in specfile library.
+
+- Batch pause button is working now.
+
+- Force a main McaWindow plot and calibration update after a fit.
+
+- Solve a Windows XP specific problem.
+ The peak selection buttons were not coloured under some windows theme configurations.
+
+
+VERSION 3.9.4 Qt4
+
+- Publication quality graphics thru matplotlib.
+ The packages are so big in part because of matplotlib.
+
+- Qt4 support. I will drop Qt3 in future releases.
+ Latest windows versions now available from sourceforge and not from the ESRF.
+
+- Universal support for Mac.
+
+- Started to add imaging tools (ESRF Data Format file stacks for the time being).
+
+- Added an RGB correlator to easily find correlations among elements.
+ This tool is to be used after a batch (Qt4 only).
+ It can be used with previous batch results.
+
+- Image printing added.
+
+- Error when working with Nd element fixed.
+
+- Incomplete legends in plots solved (Mac-i386)
+
+- Missing features:
+ Some previous features have not been ported to PyQt4 yet.
+ If you need them please, keep the older version till I add them.
+ The Physics has NOT changed. I kept the 3.9.4 to make it clear..
+
+- New tools can be accessed thru the PyMCA Tools menu.
+
+VERSION 3.9.4
+
+- Possibility to fit separately K-alpha and K-beta lines.
+
+- Possibility to use alternative L shell ratios.
+
+- Ebel's work reference corrected
+
+- Long standing bug affecting the fit of L and M lines as a group corrected.
+
+VERSION 3.9.3
+
+- Add automatic x-ray tube weighted profile generation following Ebel's formulae.
+
+- Consider scatter of more than just one excitation energy.
+
+- Take into account peaks outside the fitting region if at least one of its escape lines falls into the fitting region.
+ Particularly important for Germanium detectors where the K escape lines are far from the excitation energy.
+
+- Starting pile-up parameter changed from 1.0e-10 to 1.0e-8
+
+- The rhodium density was wrong. Changed.
+
+- Solved a problem affecting the calibration in qt 2.3.0.
+
+- Solved a program crash in qt 2.3.0 when Matrix Spectrum of a bad or non existing matrix was requested.
+
+- Solved a program crash loading the shell configuration files. The problem affected some Intel based Macs (not all).
+
+VERSION 3.9.2
+
+- It corrects a bug affecting EDF files: The ADD REMOVE REPLACE buttons were not working.
+
+- The default short tail slope is set to 0.5
+
+- The peaks tailing information was missing in the HTML Report. It should be there now.
+
+- It exposes all the theoretical shell constants in ASCII files.
+ This allows the end user to customize the theoretical values by editing the relevant files:
+ KShellRates.dat
+ KShellConstants.dat
+ LShellRates.dat
+ LShellConstants.dat
+ MShellRates.dat
+ MShellConstants.dat
+
+VERSION 3.9.1
+
+- Huge batch fitting speed up in case the user does not generate the .fit files
+
+VERSION 3.9.0
+
+- Possibility to perform linear fits. That can improve batch fitting speed.
+
+- Solved a long standing problem with the L lines of Sb, I, Cs, Ba, La, Ce, Pr and Nd.
+ Problem was due to a bad transcription of the used theoretical data.
+
+- The elements info shows energy dependent L and M ratios.
+ Nothing changes for the fit. Just the information is shown.