painful cartopy dependency removed thanks to change in Gts.show_map added Sgts.to_kml and Sgts.plot_data_sum methods. set data_xyz = None in Gts.model methods add CHANGELOG.md Start to use logging to select the verbose level. Work not finished.
corrected small bug. Sgts.sel_radius when range=[0,X] did not include provided center site This release has been distributed to various users.
- change to gts.to_pandas_df to allow xyz
- corrected date bug in gts.get_unr (not enough decimal in UNR txyz2 format for decimal year)
- accept a pck ts file as reference
- error through logging module started to be implemented
added Sgts_methods.get_unr
added Sgts_methods.apply_coseismic
added Sgts_methods.to_obs_tensor ; relax version requirements in setup.py
added Jarrin's Euler pole in Sgts.frame & nasty bug in get_coseismic which was not using the eq_date
new setup.py for docker
put obs_tensor2sgts and sgts2obs_tensor taken from pyeq directly in pyacs
refactoring of gts.lib.init.py and gts.Gts. added gts.insert_ts and update --replace in pyacs_make_time_series.py
refactoring of gts.lib.model
added pck_only option in pacs_make_time_series.py
added cvxpy as required in setup.py. Remove the png files.
add min_yaxis option in ts.plot
faultslip exploded into individual subroutines. geo2flat changed to use pyproj and web Mercator proj
added outliers package to gts.lib and a new outliers.find_l1trend function
extensive l1 trend filtering added and bug in copy in trajectory (data_xyz must be ignored)
slight change in print Euler pole estimation results
added l1_trend filter
change in Sgts.show_map
correct a bug in pyacs_make_time_series.py when there is a rename in the conf_file and uncertainties are asked
write pck now use DEFAULT_PROTOCOL instead of HIGHEST for backward compatibility
working version, plwf & prox-tv removed from requirement in setup.py
working version
refactoring version glinalg
add common mode testing status
add make_normal_system in pyacs.lib.glinalg
Sgts reorganized
handling of .data and .data_xyz is hopefully more consistent
- some refactoring. gts.read constructor. Should be compatible with jupyter notebook plotting_time_series.ipynb
- turns display off by default
- gts.plot improvement.
- turns display off by default
add seconds conversion in pyacs.lib.astrotime
- add option ref_only_all in pyacs.sol.read_conf, pyacs.sol_sinex, pyacs_make_time_series.py
- fix non optimized tsxyz approach
- pck format for Sgts added ; also added in paycs_make_time_series.py
- pyacs.gts.plot add info option , superimposed can have several gts
refactoring of gts.lib.primitive. Fix bug handling .data_xyz
- added pyacs.lib.astrotime & coordinates import in pyacs.py.
- Added pyacs.lib.utils.
- Bug corrected in pyacs.lib.shapefile when wrting polygons (due to syntax change in pyshp).
added gts.primitive.substract_ts_daily
added spherical_baseline_length_rate in pyacs.lib.coordinates
added scripts/pyacs_gvel_strain.py
Gts.lib.format.py re-organized in Gts.lib.format
reorganize filter for gts. Added total variation filters and piecewise linear fit.
in gts.extract_periods if data_xyz present then data is rebuilt and X0,Y0,Z0,lon,lat,h attributes are updated. Bug in data_xyz extraction corrected.
added pyacs_gvel_pole.py script
small potential bugs in astrotime corrected. Enforcing output type to int for mday, month, noday
pygeca separated from pyacs
script for combining psvelo velocity field added. Experimental.
minor bug in gts.remove_outliers when max index of outliers exceeds the length of the new time series
minor bug in paycs_make_time_series.py where the correlation coefficients were not properly filled by rotating the XYZ cov matrix
small change in Sgts.plot to handle the save option
change in Sgts.medvel. Now return the Sgts instance with velocity populated
small bug in gts.plot which was preventing to see the first point with date_unit='cal' option ; also dates are plot at their true time rather at the day at 0000
working version for pyeq_kinematic_inversion.py python 3.7
first Exception handling for gts. Still in test
scaling_laws and magnitude re-introduced, pole estimates, work on median & medvel in Sgts
working version for Green's function cleaning
new calc_pole method in vel_field
bug in Sgts.stat_site where ve and vn had been swaped
bug in gts.exclude_periods corrected
non linear trajectory model added to gts
Pure python Okada's formulas implementation ; bug corrected in option periods in gts.detrend_median()
Major change in pyacs_make_time_series.py that now allows coordinates uncertainties to be calculated through projection
first python 3.6 working bundle with pyacs_make_time_series and the core library ok.
- for version > 0.60, PYACS is being progressively migrated to python 3.6 & 3.7, only few features 0.51/python2.7 available for now
working code with pygeca
added observation reweighting options. InSAR capability tested.
added insar capability in pyeq_static_inversion.py (quick made from Amazonia, after World Cup final, requires more carefully result printing)
add pyeq_model_to_disp.py
- adding obs-model displacement psvelo gmt files in pyeq_kinematics.py
- occasional bug in apply_offset corrected (> dev3)
- bug in coordinates.xyz_spherical_distance (arcos missing)
- adding pyeq_scaling_laws.py
- adding pyacs/lib/GMT.py for backwards compatibility (> dev2)
- corrected sign error for up Green function using Meade tde
- detrend_median added to Gts.model
- bug gts.file -> gts.ifile (for > dev2)
- automatic shapefiles for cumulative slip and plots of time series in pyeq_kinematic.py/print_log (dev >3)
- datetime_from_calarray added to pyacs.lib.astrotime
- bug corrected in uts2hmsmicros for seconds calculation
- read_tdp added to gts.lib.format
- added itermax=10 option in Gts.lazy
- added force_day to gts to force daily solution at 1200
- change to gts.format.write_pos to force round at the second
- prototype of new detect_steps methods
- new lazy command with offset detection using a median filter implemented
- bug correction in plot when use of date and date_unit='days'
lazy_pyacs seems to be operational again. Patch in pyeq_parametrize_curve_surface_triangles.py and coordinates for GMT5 compatibility
geometry and green generation scripts included, as well as the static inversion -not fully operational yet.
new gts method in Sgts
this version has been distributed to Vergnolle/DeChabalier/Tissandier/
-cslip time dependent modeling operational again
- pyacs_qgis_model2polygon.py script added to distribution