-
New programs
- tomo_average_subtomos: This method carries out a subtomogram averaging. It means, given a set of subtomogram the program will estimate the average map
- denoising_tv: The program will make denoising using total variation
- tomo_ctf_wiener2d_correction: This method performs a CTF correction of a tilt series
- angular_continuous_assign2_gpu: Make a continuous angular assignment with GPU
- cuda_fourier_projection
- predict_deep_center
- cl2d_clustering.py
-
Programs updated
- tomogram_reconstruction: averaging with gold standard, tigre with internal interpolation, enviroment of tigre updated
- mpi_subtract_projection / subtract_projection: Improvements on projection subtraction a particle boosting both in MPI and single thread version
-
Programs fixed
- resolution_directional
- ctf_estimate_from_psd: Bug fix with no defocus
- mpi_subtract_projection / subtract_projection: subtraction a particle boosting both in MPI and single thread version
- volume_halves_restoration_gpu: Fixed failure when using masks
- Total new installer based on CMake. More robust and simplified
- New posting of statistics about installations
- Allow re-compilation without internet
- Added next generation of libstdcxx to DLTKv0.3 env to enhance compatibility
- Forcing to compile Xmipp with the libstdc++ availabe in the scipion3 env if exist
- Added support for reading EER
- Handling .gain files correctly
- Added MADV_HUGEPAGE support check
- Fix for rare axis orderings in MRC/MAP files
- Local average label added (used to measure the intensity of the particles in the original micrographs)
- New programs
- xtomo tigre reconstruction: This program provides a variety of algorithms to reconstruct tomogram from a set of tilt series
- graph_max_cut: split volume program
- tomo_simulate_tilt_series: simualtion of ts, tomograms and coordinates
- Programs updated
- Deep_center: interface simplified, calculate shift for particles
- deep_misalignment_detection: update model
- pdb_label_from_volume:Added CIF format support
- resolution_monotomo: Fast subtomo extraction
- resolution_monotomo: accelarating monotomo with float
- tomogram_reconstruction: tigre reconstruction added
- tomo_extract_subtomograms: solving some bugs and enhancing performance
- tomo_extract_particlestacks: solving some bugs and enhancing performance
- Programs fixed
- movie_estimate_gain: Initialize sigma values
- resolution_localfilter: double not integer
- movie_estimate_gain: Initialize sigma values
- project_tomography: internal loop
- pdb: Fixed PDB write function when no categories are present
- fourier_filter:Fix WEDGE option in FourierFilter
- resolution_pdb_bfactor:fix unaccounted column in res bfac pdb
- fourier_filter: Fixing missing cone
- Fixed PyXmipp installation
- Update pythorch version on DLTK pyTorch enviroment
- Fix DLTK installation
- Removed force flag on conda installations (new conda version)
- API for collecting xmipp installation statistics
- Local average added: Now the micrograph scissor annotates the local intensity value around the particles in the original micrographs.
- Added support for half precision float micrographs
- Wipe out memory ouptut in cuFFT
- Deprecated deep_denoising
- Fix for rare axis orderings in MRC/MAP files
- Adding fixed extension to tomo
- New programs
- local_volume_adjustment: This program modifies a volume in order to adjust its intensity locally to a reference volume
- subtomo_subtraction: New program with mpi version
- Programs updated
- tomo_extract_subtomogram: Accept target box size
- convert_from_pdb: Added cif support
- phantom_movies: Adding more options: circle and cross
- angular_neighbourhood: Modified to output the result in the best symmetry group
- batch_deep_misalignment: Detection: GPU option
- Programs fixed
- volume_from_pdb: Bug fix in dealing with orig=0, Detection of contrast inverted maps
- tomo_extract_particles: Stack fixes, dose fixes
- image_convert: Fixed issue when providing a metadata file as input
- volume_substraction: Change NaNs generated by a division between 0 by 0s.
- Disabled downloading 380MB file not required
- Clarifying the requirements
- More robust thread execution in the installator
- Fixed bug collecting versions
- Removed auto updater from libcifpp
- Upgraded CUDA toolkit in DLTK 1.0
- Fix PDB read: This fixes a bug where pdbs whose atom name start in position 12 (in base 0)
- CUDA 12.2 supported (not full tested)
- Fixed PDB reading for shorter PDBs
- Allow not compiling xmipp to debug changes in python scripts
- Xmipp will now be able to read any volume with any values of those fields correctly.
- Fixed pixel size not being properly stored
- Fixed MetaData to avoid dangling references
- Added compatibility for undefined axis sequences in MRC files
- Added option to write mirrors in angularDistance
- New programs
- image_peak_high_contrast(for detecting high contrast regions in tomographic reconstruction)
- misaligment_detection (to detect misalignment in tomographic reconstructions from high-contrast regions)
- deep_global_assignment
- deep_center
- Programs updated
- xmipp_angular_distance: new features
- tomo_extract_particles: new features
- subtract_projection: parallelization with mpi
- tomo_extract_subtomograms: allow downsampling of features
- angular_resolution_alignment: Detect misalignment with resolution
- Programs fixed
- align_volume_and_particles: Fixed error
- angular_project_library: Removed deterministic behaviour (mpi)
- volumen_subtraction: fixed bug
- Deprecated programs (For more details visit)
- classify_kmeans_2D
- rotational_spectra
- particle_boxsize
- New clearer, more intuitive and informative installer. It also creates a file to facilitate user support.
- Refactored the deep learning toolkit, more robust and new environment with updated tensorflow.
- Updated requirement.
- Added half precission suport to numpy
- Added the ability to read and write CIF files
- Modular design of winner filter
- Decoupling XmippTomo from XmippSPA
- Fix Zernike equation
- Removed all occurrences of non base-class default destructors
- Improved MultidimArray performance
- Added support for half precision floating point numbers in MRCs
- Assign tiff to gain files
- adding half maps labels
-
New programs: tomo_confidence_map, tomo_extract_particlestacks, tomo_extract_subtomograms, tomo_tiltseries_dose_filter, psd_estimatator
-
Deprecated programs (for more details visit the wiki): angular_distribution_show, apropos ctf_correct_idr, ctf_create_ctfdat , ctf_show , idr_xray_tomo , image_common_lines , metadata_convert_to_spider , metadata_selfile_create , mlf_refine_3d, ml_refine_3d, ml_tomo , mrc_create_metadata , pdb_construct_dictionary, pdb_restore_with_dictionary , reconstruct_admn , reconstruct_art_pseudo , resolution_ibw , resolution_ssnr , score_micrograph , reconstruct_fourier_starpu , tomo_align_tilt_series, tomo_align_dual_tilt_series, tomo_align_refinement, tomo_align_refinement, tomo_extract_subvolume, tomo_project_main, tomo_remove_fluctuations , tomo_align_tilt_series,transform_range_adjust , validation_tilt_pairs , volume_pca , volume_validate_pca , work_test , 6f4d983 , evaulate_coordinates , extract_subset , image_separate_objects , volume_enhance_contrast , volume_reslice , xray_import , xray_project , xray_psf_create , xray_reconstruct_art , gpu_correlation, gpu_utils, classify_significant, deepAlign.
-
volume_from_pdb: fixing input pdb file being overwritten when '-centerPDB' flag was set
-
xmipp_phantom_movie: adding support for fixed step shift & gain and dark image generation
-
CTF simulation allows astigmatism
-
xmipp_metadata_utility: Now join operations with an empty set will return a new empty set (previously no output file was generated).
-
xmipp_matrix_dimred: Program help improved. Exception is now thrown when the number of output dimensions is larger than the input dimensions
-
xmipp_angular_distance: Added itemId column to the output
- Refactor and simplified Readme page.
- Updating CUDA version compatibility
- Updating gcc version availables
- Fixed Matlab installation
- Added missing array include to fix compilation error with g++12
- Alert and not block compilation if gcc - CUDA are not compatible
- Avoid compilation warnings
- Required pyworkflow==3.0.31
- Maintenance: Recovered python binding tests
- Maintenance: fixing dangling pointer in xmipp_error
- Maintenance: Cleaned includes in xmipp_image_base
- PSD estimation: templating function, improving performance
- Flag cleanDeprecate in the installation; clean all deprecated executables programs
- python binding: fixed bug when Numpy arrays created by slicing were badly interpretted
- Removed "seed" library
- Fixed memory pinning CUDA bug
- Fixed compilation errors on CUDA 9
- Speeding up iterations in some xmipp programs (xmipp_ctf_group, xmipp_image_histogram, xmipp_mpi_angular_class_average, xmipp_angular_distance, xmipp_angular_estimate_tilt_axis, xmipp_ctf_create_ctfdat, xmipp_resolution_ssnr)
- New Zernike3D programs
- angular_project_library: Reported some error if there are no images in the range
- angular_discrete_assign.cpp: Removed memory leak and uninitialized values
- angular_distance: Fixing condition to avoid iteration behind the end of the MD in cases when input data have different sizes. Optimized performance
- Pdb_reduce_pseudoatoms: Produced pdb is one-based indexed
- xmipp_micrograph_automatic_picking: Fixing memory leak
- subtract_projection: Fixed several bugs (improved results), added circular mask to avoid edge artifacts, added option to boost particles instead of subtract
- Various bug fixing
- More information about hdf5 library
- Updating CUDA - GCC compatibility. Added CUDA 11.7 (not tested)
- Updating Readme
- Performance optimization (metadata binding)
- Python binding: adding methods to directly set / get entire MD row
- g++ >= 8 required
- In viewers used pwutils
- The pdb data library now has all the right fields and should write the record type ("ATOM " or "HETATM") correctly at the beginning of the line and the atomType (element) and charge (if applicable) correctly at the end of the line.
- Removal of an artifact of symmetrization related to the z pitch (symmetries.cpp)
- Using the same identical Deprecated param from pyworkflow.
- xmipp_image_operate: taked into account non existing files
- angular_continuous_assign2: Bug fixed
- volume_consensus: Bug fixed
- ctf.h and angular_continuous_assign_2: Changes for local defocus estimation #578
- Version info printed at the end of the installation
- Removed empty folder with cleanBin command
- Clarifing linking to Scipion and removed a bug with the build link
- New flag (OPENCV_VERSION) in xmipp.config
- Updated Readme (explain OpenCV-CUDA support)
- Validation server: Merged what remains
- Replaced sincos to sin and cos
- Handling of pointers in MPI programs
- "nullptr" used to denote the null pointer not "NULL"
- Check if nvidiaDriverVer is None
- Updated readme
- Updated hdf5 info troubleshoting
- Updated Standalone installation
- Updated Scons installation
- xmipp get_models: fixing the run and download path
- Updating xmipp links for Scipion on installation
- Removed fatal message in installation
- Reported error if happen on installation - runjob
- Ensuring that target directory for the libraries exists
- protocol_core_analysis: New protocol
- protocol_compare_angles: Bug fix in compare angles under some conditions
- protocol_center_particles: protocol simplified (removed setofCoordinates as output)
- protocol_CTF_consensus: concurrency error fixed
- protocol_convert_pdb: remove size if deactivated
- protocol_resolution_deepres: binary masked not stored in Extra folder and avoiding memory problems on GPUs
- protocol_add_noise: fixes
- protocol_compare_reprojections: improve computation of residuals + tests + fix + formatting
- protocol_screen_deepConsensus: multiple fixes in batch processing, trainging and streaming mode
- protocol_shift_particles: apply transform is now optional
- New XMIPP logo
- subtract_projection: adding new flag + fix
- Add intersection size metadata (bindings/python)
- Fixed unitialized unique pointers (bindings/python)
- Bug fixing: Resolution directional and anisotropic filtering fixing the test
- Removed SonarCloud issues
- Replaced defines with constexpr
- Removing Unused funtion parameters
- Division by zero
- Memory management
- Removed field shadowing
- Destructors should not throw exceptions
- Updating to C++17
- Support newer versions of CUDA and gcc
- Zernike programs compatible with Cuda 8.x
- Fixed Sonar Cloud issues and bugs
- Matlab compilation Fixed
- Fixed importing pwem.metadata
- nma_alignment: Fixed arguments for the xmipp_angular_projection_matching invocation
- Fixed test fails: ResolutionSsnr, ReconstructArtMpi, ReconstructArt, MlfRefine3dMpi, MlfRefine3d, MlRefine3dMpi, MlRefine3d, xmipp_test_pocs_main & volume_subtraction
- xmipp_micrograph_automatic_picking: Fixed tests, avoid possible memory corruption
- resolution_pdb_bfactor: bug fixed - error with multiple chains
- FlexAlign: Fixed crash when binning > 1
- Bug fixed and allowed controlling high sampling rate
- Volume consensus: Fixed number of levels in the wavelet transform
- Compilation: Fixed compilation of starpu programs
- xmipp_transform_dimred: Fixed output metadata in append mode, adding MDL_DIMRED label
- Config file generation: Fixed config version detection outside of the git repo, refactored check_CUDA and managed gcc compiler if it is installed out of /usr/bin/, check and exit if xmipp.conf does not exist
- Compilation: Fixed detection of the last commit changed the config script
- Resolution_fso: Bingham test implemented
- Opencv not detected. Added include to user/include/opencv4 folder on config file
- Compilation: asking whether to continue with compilation even though the config file is outdated
- XMIPP install: Linked libsvm to scipion
- Installation: Referenced 'global' xmipp.conf instead of using local copy of it
- Multiple MPI programs: replaced CREATE_MPI_METADATA_PROGRAM macro by templated class
- python_constants: add defocus labels
- Metadata: added new nmaEigenval label
- Python binding: added new function - correlationAfterAlignment, MDL_RESOLUTION_ANISOTROPY, MDL_RESOLUTION_ANISOTROPY
- Matlab binding dependencies: set XMIPP as a hard dependency
- Projections subtraction: new program
- FFTwT: added mutex for plan handling
- Multiple programs: Added a common implementation of the rerun
- Phantom_create: update info link
- Multiple programs: Added a common implementation of the rerun
- Transform Geometry: save new shifted coordinates in option "shift to" + enterOfMass to python binding
- Readme info: add virtual machine info
- Removal of the SVM from inside the XMIPP repository and downloading it as an external dependence
- Solved a configuration problem with CUDA
- ml_tomo: Using .mrc instead of .vol ; volume_align: Addded wrapping during alignment
- CUDA-11 support
- New protocol: Deep align
- ChimeraX support
- Improvements of streaming process
- Several performance optimizations
- Build time optimization
- Multiple bug fixes
- Improved documentation
- Fast CTF estimation
- CTF includes phase shifts now
- Selection of alpha helices or beta sheets from a PDB (xmipp_pdb_select)
- Centering a PDB (xmipp_pdb_center)
- New Protocol: MicrographCleaner is a new algorithm that removes coordinates picked from carbon edges, aggregations, ice crystals and other contaminations
- New functionality: The protocol compare reprojections can now compute the residuals after alignment
- New protocol: Split frames divide input movies into odd and even movies so that they can be processed independently
- New protocol: Continuous heterogeneity analysis using spherical harmonics (not ready to be used)
- Bug fixing when some micrograph has no coordinates in the consensus-picking.
- New functionalities: Different architectures and training modes
- Normal Mode Analysis protocols have been moved to the plugin ContinuousFlex
- Fixing MPI version of the Fourier Reconstruction
- New protocol: local CTF integration and consensus protocol for local ctf (also the viewers)
- Local CTF analysis tools: Not yet ready for general public
- New functionallity: Introducing the posibility of automatic estimation of the gain orientation.
- Bugs fixings regarding stability on streaming processing
- Support of heterogeneous movie sets
- New protocol: Clustering of subtomogram coordinates into connected components that can be processed independently
- New Protocol: Removing duplicated coordinates
- New protocol: Subtomograms can be projected in several ways to 2D images so that 2D clustering tools can be used
- New protocol: Regions of Interest can be defined in tomograms (e.g., membranes)
- Bug fixing in mask3d protocol
- Bug fix: in helical search symmetry protocol
- Enhanced precision of the FlexAlign program
- Now, deepLearningToolkit is under its own conda environment
- Multiple protocols accelerated using GPU
- New functionality: Xmipp CTF estimation can now take a previous defocus and do not change it
- New functionallity: CTF-consensus is able to take the primary main values or an average of the two.
- New functionallity: CTF-consensus is able to append metadata from the secondary input
- New functionality: Xmipp Highres can now work with non-phase flipped images
- New functionality: Xmipp Preprocess particles can now phase flip the images
- New protocol: Tool to evaluate the quality of a map-model fitting
- Allowing multi-GPU processing using FlexAlign
- Improvement in monores and localdeblur
- Randomize phases also available for images
- Change the plugin to the new Scipion structure
- Migrating the code to python3