utils: refactor time_func.py and add s1_utils.py #661
Conversation
+ add utils.time_func.py:
- move get_design_matrix4time_func() from objects.stack.py.
- switch module import/usage in all related scripts: dem_error, gps, simulation, tsview
- replace the old func definition with a pointer to the utils.time_func.py for backward compatibility.
- move estimate_time_func() from timeseries2velocity.py
- switch module import/usage in all related scripts: timeseries2velocity, tsview
- set all get_design_matrix4*_func() as independent functions for more flexibility
+ utils.s1_utils:
- add estimate_S1AB_bias()
- add get_subswath_masks()
+ objects/gps.py: check site name during initialization for more informative error msg against typo in "view.py --ref-gps"
+ docs/api/module_hierarchy.md: add time_func and s1_utils.
This commit adds the capability to feed accurate hour/minute/second info of step/exp/log functions to the design matrix, to support scenarios when the event and SAR acquisition are in the same day, such as The M6.4 Ridgecrest EQ in 4 July 2019. + ptime.yyyymmdd2years(): add seconds arg to add the time of the day info to the output list of years in float + time_func: add seconds arg to pass to ptime.yyyymmdd2years() to support time of the day info to all functions inside + read/prep seconds info to feed to time_func.py in the following scripts: - dem_error.py - timeseries2velocity.py - tsview.py + update example usages in smallbaselineApp.cfg file and arg_group.py
+ add RidgecrestSenDT71 as example dataset for HyP3 - docs/demo*.md: add RidgecrestSenDT71 dataset produced by ASF HyP3 service by Jiang Zhu. - mintpy/data/input_files: add RidgecrestSenDT71.txt + add RidgecrestSenDT71 as test dataset for hyp3 - tests/test_smallbaselineApp: add url for RidgecrestSenDT71 prepared using ASF HyP3 by Jiang Zhu. - tests/configs/RidgecrestSenDT71.cfg: update setups for faster testing + readfile.read_gdal(): use numpy indexing instead of gdal python binding for more robust performance + plot.auto_figure_title(): use custom ds name for generic h5 file
|
@yuankailiu could you please take a look at the Note that I added a pointer in the stack.py to the new time_func.py for backward compatibility, so I don't expect impacts on downstream code from users [would recommend switching to the new script though as it's simpler]. |
|
For book keeping, below are the estimated step function for the co-seismic deformation from the two EQs and it's displacement time-series of a pixel near the fault.
|
|
Thank you @yunjunz for making this much cleaner, even with the resolution down to seconds! |
There was a problem hiding this comment.
They look good to me. I did not modify anything.
[Update]
Below is tested with the postseismic exponential function (at a pixel) from the Ridgecrest EQ.
(Time-series files are provided by @nabolfat)
Description of proposed changes
This PR adds the following new features:
utils/time_func.pyforestimate_time_func()(from timeseries2velocity.py) andget_design_matrix4time_func()(from objects/stack.py and all its sub-functions).THH(:)MMstyle for relevant time functions (step/exp/log) for hyp3: fix incAngle/slantRangeDist + add more attributes (pass full test) #640.utils/s1_utils.pydedicated for Sentinel-1 specific functionalities, with initial content from Yunjun et al. [in prep]RidgecrestSenDT71forhyp3from @cirrusasf (sbApp_hyp3: update url of the example dataset MintPy-tutorial#26).Reminders