v0.1.0rc3

What's Changed

• fix unpickle_bytes_columns crash on variable-size binary columns by @TeunHuijben in #273
• Enabling type casting on CTC export by @JoOkuma in #275
• Optimization by @TeunHuijben in #274
• Speedup cache schema bulk signal read by @TeunHuijben in #276
• handle painting when bbox on/outside buffer by @TeunHuijben in #277
• Fix add bbox to node update signal by @TeunHuijben in #278
• Fixing array type casting by @JoOkuma in #279
• Adding shift division by @JoOkuma in #280
• Saving dtypes as metadata for roundtrip consistency by @yfukai in #262

Full Changelog: v0.1.0rc2...v0.1.0rc3

v0.1.0rc2

What's Changed

• Adding float16 type conversion by @JoOkuma in #257
• Conversion of float 16 when saving geff by @JoOkuma in #258
• Fixing successors large query by @JoOkuma in #250
• Assigning track IDs allowing frame skip by @yfukai in #261
• Adding update_node signal and connected them to spatial filters by @yfukai in #266
• Dict-like metadata interface & private metadata by @yfukai in #260
• Fixing regions props node attribute init by @JoOkuma in #259
• Notifying GraphArrayView of updates by @yfukai in #267
• small bug fix by @ThibautGoldsborough in #271
• graphview.bulk_add_edges actually uses ._root.bulk_add_edges by @TeunHuijben in #272

Full Changelog: v0.1.0rc1...v0.1.0rc2

v0.1.0rc1

What's Changed

• updating badge and classifiers by @JoOkuma in #168
• Updating install instructions by @JoOkuma in #169
• Partial update of track IDs by @yfukai in #158
• Returning geff metadata and adding geff_read_kwargs by @JoOkuma in #172
• Fixing odd shape mask cropping by @JoOkuma in #174
• Fixing saving nd-array on geff by @JoOkuma in #170
• Making attr_key required for GraphArrayView by @yfukai in #176
• Test all backends for GraphArrayView by @yfukai in #177
• Adding is_in operator for AttrComparison by @yfukai in #178
• Renaming "track_id" to "tracklet_id" by @yfukai in #179
• Returning previous and updated tracklet IDs from assign_tracklet_ids. by @yfukai in #180
• Implementing union operation for mask by @yfukai in #181
• Improving match visualization by @JoOkuma in #184
• Fixing bbox default value typing by @JoOkuma in #186
• Adding mask __isub__, move, and dilation` by @JoOkuma in #185
• Apply tiled functionality for tiled processing of graphs by @JoOkuma in #183
• Using np.nextafter instead of eps by @yfukai in #189
• Fix from_others issue with SQLgraph. by @yfukai in #182
• Ensuring graph array view implements LayerDataProtocol for napari by @yfukai in #198
• Graph metadata API by @JoOkuma in #197
• Changed successors and predecessors so that they return IDs. by @yfukai in #188
• Fixing required shape parameter of to_ctc graph method by @JoOkuma in #201
• Fixing reset of ctc metrics eval by @JoOkuma in #203
• Retrieving pickle dtype to avoid sqlgraph mask update error by @yfukai in #200
• Adding method to remove and add nodes to bbox spatial filter by @JoOkuma in #159
• Adding ancestral_connected_edges by @JoOkuma in #209
• Optimizing SQLGraph insert and update by @JoOkuma in #208
• Add traccuracy support by @JoOkuma in #213
• Fixing tracklet graph by @JoOkuma in #214
• Fixing _update_max_id_per_time for SQLGraph by @yfukai in #211
• Mask.regionprops property to calculate properties. by @yfukai in #210
• Fix dtype issue with bbox in SQLGraph by @yfukai in #212
• Fixing double t's on sqlgraph view attributes by @JoOkuma in #218
• Improving SQLGraph _get_neighbors performance by @JoOkuma in #222
• Fix benchmark (and migrate to ASV) by @yfukai in #199
• Caching spatial_filter by @yfukai in #215
• Allows node_ids=None for successors and predecessors by @yfukai in #225
• Adding merge to ILPSolver by @JoOkuma in #171
• Removing @Property from Graph API by @JoOkuma in #226
• remove_attr_key by @yfukai in #224
• allow graphView to be pickled and unpickled by @TeunHuijben in #232
• Adding edge_list function by @JoOkuma in #236
• adding has node functionality and simplifying has edge by @JoOkuma in #235
• Adding arguments to rename keys in 'from_geff' by @JoOkuma in #239
• Adding 'z', 'y', 'x' constants and other default attribute fixes by @JoOkuma in #238
• create_node|edge_attr_index for SQLGraph by @yfukai in #223
• Making matching more robust and fixing random matching by @JoOkuma in #240
• distance based graph matching by @ThibautGoldsborough in #237
• Fixing metadata from_other by @JoOkuma in #242
• added spacing parameter to regionprops by @TeunHuijben in #245
• Adding neighbors_per_frame option to DistanceEdges by @JoOkuma in #216
• Fixing mask dtype bug by @JoOkuma in #247
• Refactoring node and edge attribute key API by @JoOkuma in #244
• Fixing track id key creation by @JoOkuma in #254
• add overwrite option to .to_geff() by @TeunHuijben in #256
• Fixing loading of matching weights by @JoOkuma in #255
• Adding kwargs to Mask.regionprops by @JoOkuma in #253
• Delayed default value setting by @JoOkuma in #246

New Contributors

• @ThibautGoldsborough made their first contribution in #237

Full Changelog: v0.1.0rc0...v0.1.0rc1

v0.1.0rc0

What's Changed

• Initial implementation by @JoOkuma in #1
• Updating line length by @JoOkuma in #10
• Adding simple ASV benchmark and CI by @JoOkuma in #11
• Implementing subgraph operations and id remapping by @JoOkuma in #5
• Improving test coverage by @JoOkuma in #12
• Improving API standardization by @JoOkuma in #14
• Fixing edge features return values by @JoOkuma in #15
• Improving slicing abstraction with GraphView by @JoOkuma in #16
• Function to add tracklets id to graph by @JoOkuma in #9
• Adding SQLAlchemy backend by @JoOkuma in #18
• initial ILP implementation by @JoOkuma in #19
• Adding intermediate import by @JoOkuma in #23
• Adding data loading from CTC by @JoOkuma in #26
• Compatibility with py-ctcmetrics by @JoOkuma in #28
• Fixing running both solvers by @JoOkuma in #30
• adding unpack option to node|edge_features by @JoOkuma in #31
• Add degree function by @JoOkuma in #32
• Adding graph matching by @JoOkuma in #33
• Keeping node_ids when features=None by @JoOkuma in #36
• Adding default value for BaseGraph.in{out}_degree by @JoOkuma in #37
• Fixing duplicated keys {node|edge}_feature_keys by @JoOkuma in #41
• Enabling infinity expressions by @JoOkuma in #40
• Nitpicking on lambda usage by @JoOkuma in #39
• Adding edge_ids method by @JoOkuma in #45
• Keeping features sorted by @JoOkuma in #46
• Everything is attribute by @JoOkuma in #47
• Update benchmarking by @JoOkuma in #48
• Fixing matching bug by @JoOkuma in #49
• Adding solutions and tracks ids automatically reset by default by @JoOkuma in #53
• Allowing n-cell division by @JoOkuma in #55
• Adding errors for solving empty graph and returning graph by default by @JoOkuma in #56
• Adding mask repr and size cache by @JoOkuma in #57
• Avoiding adding solution key to node twice by @ilan-theodoro in #59
• Fixing evaluate metrics not assigning new track ids by @JoOkuma in #60
• Fix with_columns usage according to polars 1.31 by @JoOkuma in #62
• Attribute rework by @JoOkuma in #58
• switching to bsd-3 clause license by @JoOkuma in #64
• Crop attrs functionality by @JoOkuma in #65
• Micro optimization of benchmarking by @JoOkuma in #66
• adding ILP gap parameter by @JoOkuma in #68
• Optimizing track id assignment by @JoOkuma in #69
• Initial documentation by @JoOkuma in #63
• Fixing readme and benchmark by @JoOkuma in #71
• Adding node contractions by @JoOkuma in #72
• Adding variable delta_t to distance edges by @JoOkuma in #73
• Adding overlap support by @JoOkuma in #75
• Graph loading from numpy array by @JoOkuma in #76
• Automating release by @JoOkuma in #77
• Refactoring global Options for show_progress by @JoOkuma in #79
• Multi processing option by @JoOkuma in #80
• Batched node generic processing by @JoOkuma in #82
• adding new authors by @JoOkuma in #85
• Improving add bulk operations by @JoOkuma in #86
• adding BaseGraph.from_other and GraphView.detach by @JoOkuma in #87
• Exposing submodules to main namespace by @JoOkuma in #88
• Bug fix of adding attribute of an existing key by @JoOkuma in #89
• Setting blosc2 threads to one, faster for small masks by @JoOkuma in #90
• Adding overlap computation helper function by @JoOkuma in #91
• Clean up and new suggested usage by @JoOkuma in #92
• Fixing matching bug by @JoOkuma in #93
• Renaming vars from PR #93 by @JoOkuma in #94
• SQLGraph subgraph query improvements and minor fixes by @JoOkuma in #97
• Removing radius from load_array by @JoOkuma in #99
• Adding summary function by @JoOkuma in #100
• fixing CTC export by creating new track id for long edges by @JoOkuma in #98
• Adding to_ctc function by @JoOkuma in #101
• Fixing usage of polars slice API by @JoOkuma in #102
• Fixing empty frames matching by @JoOkuma in #103
• New indexed graph backend by @JoOkuma in #95
• Updating to_napari_format docs by @JoOkuma in #104
• Adding time out to ILP by @JoOkuma in #107
• Adding timeout docs by @JoOkuma in #108
• Adding filter syntax by @JoOkuma in #96
• Fixing asymetric division bug by @JoOkuma in #109
• Optimizing GraphView by @JoOkuma in #110
• Merge dev with main by @JoOkuma in #105
• Adding spatial graph support by @JoOkuma in #112
• Adding offset to assign_track_ids and new tracklet_graph for napari by @JoOkuma in #116
• Adding bounding-box attribute by @JoOkuma in #121
• Initializating rustworkx graph backends from rx_graph by @JoOkuma in #123
• Adding boundary box spatial filter by @yfukai in #122
• Updating GraphArrayView for better compatibility with NumPy-array by @yfukai in #115
• Fixing IndexedRXGraph.node_attrs and attr_keys of Filters by @JoOkuma in #127
• Make drawing faster by updating drawing for GraphArrayView by @yfukai in #128
• Adding missing track_id_offset by @JoOkuma in #132
• Adding geff support by @JoOkuma in #125
• fix time name bug by @JoOkuma in #134
• Updating geff dependencies to use source by @JoOkuma in #138
• Adding method to remove node by @JoOkuma in #135
• Adding custom indices support to SQL graph by @JoOkuma in #139
• Fixing indexed graph add node by @JoOkuma in #136
• Adding edge utilities by @JoOkuma in #140
• Fix assign_track_ids function for the graph by @yfukai in #137
• Fix node id map empty graph by @TeunHuijben in #142
• fixed bug in graphView.has_edge + precommit by @TeunHuijben in #141
• Adding rx from pypi by @JoOkuma in #143
• Making GraphView.detach indexed by @JoOkuma in #145
• Filtering test warnings by @JoOkuma in #146
• Improving single node API by @JoOkuma in #129
• Copy method by @TeunHuijben in #144
• Fixing edge case where table is empty by @JoOkuma in #148
• Fix track_id_offset bug by @yfukai in https://githu...