Releases: jungmannlab/picasso
v0.6.9
Picasso 0.6.9 is here.
Changelog
- Polygonal pick added to Render -> draw any shape to pick localizations
- Save pick properties saves pick region's area in nm^2
- 3D calibration .yaml file saves the number of frames in the movie and the step size
Notes
v0.6.8
Picasso 0.6.8 is here.
Changelog
- .tiff files can be loaded to Localize directly, although the support may be limitted
- Bug fix: 2D cluster centers area and convex hull are saved correctly
- Bug fix: rectangular picks
Notes
v0.6.7
Picasso 0.6.7 is here.
Changelog
- GUI modules display the Picasso version number in the title bar
- Added readthedocs requirements file (only for developers)
- No blur applied when padding in Picasso: Render (increases speed of rendering)
- Camera settings saved in the .yaml file after localization
- Picasso: Design has the speed optimized extension sequences (Strauss and Jungmann, Nature Methods, 2020)
- Change matplotlib backend for macOS (bug fix with some plots being unavailable)
- Build animation does not trigger antivirus, which could delete Picasso (one click installer only)
Notes
v0.6.5
Picasso 0.6.5 is here.
Changelog
Many bug fixes for the one-click-installer version:
- Fix save pick properties bug in Picasso Render GUI (one click installer)
- Fix render render properties bug in Picasso Render GUI (one click installer)
- Fix animation building in Picasso Render GUI (one click installer)
- Fix test clusterer HDBSCAN bug
- Fix .nd2 localized files info loading (full loader changed to unsafe loader)
Fix rare bug with pick similar zero division error
Notes
v0.6.4
Picasso 0.6.4 is here.
Changelog
- Fix RCC error in one click installer (in versions 0.6.1-3)
- Picasso runs on Python 3.10 (jump from Python 3.7-3.8)
- New installation instructions
- Dependencies updated, meaning that M1 should have no problems with old versions of SciPy, etc.
- Localize: arbitrary number of sensitivity categories
- Picasso Render legend displays larger font
- Picasso Render Test Clusterer displays info when no clusters found instead of throwing an error
- Calling clustering functions from
picasso.clustererdoes not require camera pixel size. Same applies for the corresponding functions in CMD. Only if 3D localizations are used, the pixel size must be provided. - HDBSCAN is installed by default since it is distributed within the new version of
scikit-learn 1.3.0 - Screenshot
.yamlfile contains the list of colors used in the current rendering - Render scale bar allows only integer values (i.e., no decimals)
- Localize .ims file fitting bug solve
- Measuring in 3D render window (Measure tool and scalebar) fixed (previous versions showed incorrect values)
- Localize GUI allows for numerical ROI input (Parameters dialog)
- Export current view in Render saves metadata; an extra image is saved with a scalebar
- Allow loading individual .tif files (same as prior to v0.4.11)
- Clustering using CMD requires the user to input camera pixel size (instead of asking for the input after the function call)
- RESI localizations have new column 'cluster_id'
- Bug fixes
Notes
v0.6.0
Picasso 0.6.0 is here.
Changelog
- RESI dialog added to Picasso Render, allowing for substatial boost in spatial resolution (to be published)
- Quantum efficiency of a camera is not used anymore to convert raw data into photons. This affects localization precision calculated, such that it is closer to the experimental value obtained using NeNA
- Localize CMD: new argument (-r, --roi) for specifying ROI
v0.5.7
Picasso 0.5.7 is here.
Changelog
- Render group information is faster (e.g., clustered data)
- Test Clusterer window (Render) has multiple updates, e.g., different projections, cluster centers display
- Cluster centers contain info about std in x,y and z
- If localization precision in z-axis is provided, it will be rendered when using
Individual localization precisionandIndividual localization precision (iso). NOTE: the column must be namedlpzand have the same units aslpxandlpy. - Number of CPU cores used in multiprocessing limited at 60
- Bug fixes
Notes
v0.5.6
Picasso 0.5.6 is here.
Changelog
- Cluster centers contains analogous columns to the ones generated before the 0.5.x releases
- Clearer readme (installation instructions)
- Numerous bug fixes (custom colormap, robust .nd2 reading, close file, etc
- One click installer should not require administrator permissions (due to antivirus protection)
Notes
v0.5.4
Picasso 0.5.4 is here.
Changelog
- 3D rendering rotation window
- Multiple .hdf5 files can be loaded when using File->Open
- Localizations can be combined when saving
- Render window restart (Remove all localizations)
- Multiple pyplot colormaps available in Render
- View->Files in Render substantially changed (many new colors, close button works, etc)
- Changing Render's FOV with W, A, S and D
- Render's FOV can be numerically changed, saved and loaded in View->Info
- Pick similar is much faster
- Remove localization in picks
- Fast rendering (display a fraction of localizations)
- .txt file with drift can be applied to localizations in Render
- New clustering algorithm (SMLM clusterer)
- Test clusterer window in Render
- Option to calculate cluster centers
- Nearest neighbor analysis in Render
- Numerical filter in Filter
- New file format in Localize - .nd2
- Localize can read NDTiffStack.tif files
- Docstrings for Render
- Plugins can be added to all Picasso modules
- Sensitivity is a float number in Server: Watcher
- Many other improvements, bug fixes, etc.
Notes
v.0.4.11
Picasso 0.4.11 is here.
Changelog
- More documentation for the watcher.
- The watcher of Picasso Server can automatically process different settings, see here.
- The watcher now logs what it is doing in the background.
- Fixed installation for Mac OS
- Small optimizations and bugfixes
⚠️ When wanting to process.ims-files with the Watcher, you will need to launch the server module as an administrator⚠️ Performance warning: Processing large files (e.g. > 10k frames), running multiple watchers simultaneously, and having multiple processing groups on PCs with limited memory could lead to issues, so it is recommended to only run one watcher at a time in these cases.- 📽️ An updated quickstart video can be found here
Notes