v0.2.4

What's Changed

• Support numpy>2.0 in #4
• Fix MATLAB video reader to work with files from other labs
• load_metadata: Include number of frames/image size for MultiRecorder imports
• Interactive segmentation: Add largest connected component filter, update icons
• The monochrome PyPi package was renamed to monochrome from monochrome-viewer
• Fix video looping in documentation
• Update documentation

Full Changelog: v0.2.3...v0.2.4

v0.2.3

What's Changed

• Migrated docs from readthedocs to GitHub pages (due to memory limits). Automatic docs builds are now much faster!
• Updated and improved docs and tutorials
• Added support for python 3.13
• Added video importer for .mp4/.avi/.mov/...
• Added export_image() function in addition to save_image(). export_image() will always convert to uint8 for best compatibility with image viewers.
• save_image(): added compat parameter to convert image to 8bit
• Renamed export_video_collage() to export_videos(), now included in main namespace
• Fixed issue with numpy>=2.0, require numpy<2.0 as some dependencies rely on it for now
• Fixed videoplayer out of bounds
• Fixed export_videos() cmap and alpha clipping
• Fixed animation reset after saving to mp4 bug
• Fixed compare_traces() for videos with different lengths
• Removed WIP adp module to feature branch
• Fixed documentation builds for sphinx >= 7.4.6

Full Changelog: v0.2.2...v0.2.3

v0.2.2

What's Changed

• Renamed video.play* functions to start with show_video, old functions are deprecated
• Renamed retrieve_example_data() to download_example_data(), added to top-level import. Old function name is deprecated.
• Renamed example data file names, added dual channel example. Old file names are deprecated.
• Changed order of arguments for video.save_image_sequence, path is now first arg. Old convention is deprecated.
• Change order of arguments for show_positions, positions are now first argument
• video.normalize: adedd dtype argument, infer maximum for integer output dtypes, support complex numbers
• Temporal difference filter: added fill_value and center arguments
• Contrast enhancement: support mask of valid values
• Added kwargs to load_video() for plugins
• InteractivePlayer.save(): added hide_slider/buttons/framecounter params
• Fixed export_video() for odd-sized videos
• Fixed print_properties() for videos with NaNs

Full Changelog: v0.2.0...v0.2.2

v0.2.0: Interactive masking and more

What's Changed

Interactive masking editor and more:

interactive_mask.mp4

Breaking changes:

• om.video.resize: parameters are now shape or scale, instead of newx, newy
• All save and export functions now expect the filename first, e.g. save(filename, value, ...) instead of save(value, filename, ...). The deprecated save(value, filename, ...) order will raise a warning but still work.
• om.load_video: Raises FileNotFoundError instead of ValueError if file not found
• MacOS 12+ is now required

New functionality & other changes

Python 3.12 is now supported.

optimap.image:

• Added interactive_mask(): Create a mask interactively by drawing on an image.
• Added load_mask() function
• show_mask(): only show True values in overlay. If no image passed show a black border around mask. Added alpha, color, cmap parameters
• Fixed load_mask()
• Added collage() function to arrange list of images in a grid
• Added resize(), rotate_left(), rotate_right(), flip_left_right(), flip_up_down(), crop(), pad() functions analogous to the om.video equivalents
• Fixed smooth_gaussian() for RGB(A) images

optimap.video:

• Added om.video.export_video_collage() to arrange a list of videos in a grid and save to a media file
• om.video.load_image_folder() function is now public API in addition to the load_video() wrapper
• om.video.InteractivePlayer class is now public API
• om.video.MultiRecorderImporter: added get_frametimes() function to load the frametimings (if supported by MultiRecorder camera plugin)
• export_video(), export_video_with_overlay(): add progress_bar parameter
• resize: add interpolation parameter for the different OpenCV interpolation methods

optimap.trace:

• show_positions(): add size, cmap, vmin, vmax and kwargs parameters
• Fixed random_positions()
• Add collage_positions() as equivalent to image.collage() or video.export_video_collage()

Full Changelog: v0.1.4...v0.2.0

v0.1.4

What's Changed

Breaking Changes

• compute_activation_map: returns time in milliseconds instead of seconds for computed activation times when fps parameter is specified

New Functionality

• Added smooth_gaussian, erode_mask, dilate_mask, binary_closing, binary_opening functions to image module
• image.show_image: added show_colorbar and colorbar_title parameters
• image.show_mask added title parameter
• compute_activation_map: added show (for plotting, True by default), inverted, and set_nan_for_inactive parameters

Other Changes

• Ignore NaNs for play_video() and other functions
• Added support for Python 3.12
• Improved error handling for MultiRecorderImporter

Full Changelog: v0.1.3...v0.1.4

v0.1.3

• Add minmal GUI to om.play_video video.play/play2/playn
• Fix om.load_video() when using a combination of step and frames parameters
• Add vmin1, vmax1, vmin2, vmax2 arguments to video.play2
• video.normalize: add vmin and vmax paramters
• Fix overlay video code when overlay contains NaNs

v0.1.2

• Switch from QtCairo to QtAgg for the default interactive matplotlib backend. Simplifies installation as we don't depend on cairo anymore.
• fix select_traces()
• fix load_tiff_folder()
• fix error handling when OpenCV is not installed
• fix save_image_sequence for tiff files
• add fps and x parameter to select_traces() and compare_traces()
• added WIP ratiometry tutorial example

v0.1.1

Fixes issue with example file download.

v0.1.0

Initial release of optimap.