v0.35.1

What's Changed

• filtering.py: Fix in_bounds_pixels masking, set default to 25 km by @seyeonjeon in #527
• Set output_reference_idx separately from compressed_reference_idx by @scottstanie in #528

Full Changelog: v0.35.0...v0.35.1

v0.35.0

What's Changed

• Fix grammar on VRT size check by @scottstanie in #516
• Use L1 inversion by default for timeseries by @scottstanie in #518
• Add keepdims options to reader classes to avoid squeezing singleton dims by @scottstanie in #521
• Use the most recent compressed SLC as output for LAST_PER_MINISTACK by @scottstanie in #513
• simulate.py: Add use_seasonal_coherence, remove numba by @scottstanie in #524
• Adapt short wavelength filter to use gdal_fillnodata to mitigate edge effects by @scottstanie in #520
• Compute residuals to the timeseries inversion, save to raster outputs by @scottstanie in #523

Full Changelog: v0.34.2...v0.35.0

v0.34.2

What's Changed

• 'Use nearest resampling for merging temporal coherence by @scottstanie in #508

Full Changelog: v0.34.1...v0.34.2

v0.34.1

What's Changed

• Use fspath when opening with gdal to support older versions by @scottstanie in #502

Full Changelog: v0.34.0...v0.34.1

v0.34.0

What's Changed

• Pin mkdocs-jupyter to 0.25.0, small timeseries fix for cli by @scottstanie in #486
• Add missing annotations to make Python 3.9 compatible by @scottstanie in #487
• Delete .flake8 by @scottstanie in #489
• Fix unit string written to velocity.tif by @scottstanie in #491
• Add --use-evd to config cli by @scottstanie in #492
• Change default ministack_size to 15 in PhaseLinkingOptions by @scottstanie in #493
• Add logging redirect, request spurt log file, multiprocess interpolation by @scottstanie in #495
• Use nearest sampling to stitch static layers by @scottstanie in #496
• Fix extra_reference_date logic for single reference unwrapping by @scottstanie in #497
• Speed up the post-spurt 2-pi ambiguity interpolation by @scottstanie in #499
• Add CITATION.cff for better zenodo parsing by @scottstanie in #488

Full Changelog: v0.33.0...v0.34.0

v0.33.0

What's Changed

• Add fill_value to filter_long_wavelength, apply to masked pixels if provided by @scottstanie in #483
• Avoid capturing all logging from spurt by @scottstanie in #484
• Add create_nonzero_conncomp_counts to create counts of valid unwrapped outputs by @scottstanie in #485

Full Changelog: v0.32.0...v0.33.0

v0.32.0

What's Changed

• Merge documentation improvements from joss branch by @scottstanie in #481
• Add a retry to spurt for BrokenProcessPools by @scottstanie in #480

Full Changelog: v0.31...v0.32.0

v0.31

What's Changed

• Skip nodata values of interferograms when interpolating by @scottstanie in #476

Full Changelog: v0.30...v0.31

v0.30

What's Changed

• Pass through units during timeseries._redo_reference by @scottstanie in #451
• Fix the relative index used to make compressed SLCs by @scottstanie in #452
• Fix in_trim_slice size for non overlapping blocks by @scottstanie in #454
• Add ability to specify output bounds as WKT by @scottstanie in #427
• Refactor compressed_reference_idx, fix relative index bug by @scottstanie in #453
• Call spurt from subprocess to avoid fork issues by @scottstanie in #455
• Replace CSVs with chi2.ppf for GLRT test by @scottstanie in #456
• Add ability to calculate similarity using nearest-3 interferograms by @scottstanie in #457
• Pass through compressed_slc_plan config to sequential by @scottstanie in #459
• Add single_tile_reoptimize for snaphu-py by @scottstanie in #460
• Update dolphin timeseries cli for new options, use L1 by default by @scottstanie in #461
• Add a layover_shadow_mask_files to mask pixels in layover/shadow during wrapped_phase by @scottstanie in #462
• Add min_conncomp_frac, lower default to 0.001 by @scottstanie in #463
• Increase spurt`` max_tiles default to 49 to have smaller MCF problems by @scottstanie in #464
• Use output_reference_idx as the default for creating compressed slcs with ALWAYS_FIRST by @scottstanie in #465
• Allow interpolation using pseudo-correlation, lower spurt tile sizes by @scottstanie in #466
• Add zero_correlation_threshold phase linking parameter to zero out low correlation by @scottstanie in #467
• Pass through block_shape to create_similarities after phase linking by @scottstanie in #469
• Increase the buffer_pixels when making OPERA CSLC mask by @scottstanie in #470
• Pass through --max-tiles to spurt by @scottstanie in #471
• Refactor repack_raster to work in block for lower memory by @scottstanie in #473
• Setup conversion to /vsis3 when reading from S3 using osgeo.gdal by @scottstanie in #474
• Add censored_lstsq to solve least squares with missing data by @scottstanie in #475

Full Changelog: v0.29.0...v0.30

v0.29.0

What's Changed

• Add references/url fix from JOSS branch by @scottstanie in #439
• Add a sample dolphin_config.yaml to the docs, GHA to auto-update by @scottstanie in #438
• Pass through phase_linking.output_reference_idx to avoid reference resets by @scottstanie in #445
• Set the input to EVD to be weighted by correlation by @scottstanie in #428
• Add stitched phase cosine similarity raster to outputs by @scottstanie in #446
• Add logic to reset the extra_reference_date after a network unwrapping by @scottstanie in #442
• Add a post processing step to fill nans after running spurt by @scottstanie in #448
• Apply nan/0 mask to output of goldstein filter by @scottstanie in #447
• Use dataclasses instead of NamedTuple for displacement, stitched outputs by @scottstanie in #449
• Add outside_input_range options to get_nearest_date_idx by @scottstanie in #450

Full Changelog: v0.28.1...v0.29.0