repository-preservation / lcmap-pyccd Goto Github PK

Hi,
I followed the installation guidelines and installed device and test dependencies, but got a red warning during the last part saying that 'pytest-profiling 1.6.0 has requirement pytest<4.0.0, but you'll have pytest 4.3.0 which is incompatible.' Then I ran the pytest and got below error. I also tried to run the Python Notebook but it failed to load module aniso8601, while it was installed in the site-packages. Could you please help me resolve this issue?

Traceback (most recent call last):
  File "c:\users\shhey\pyccd\.venv\lib\site-packages\pluggy\manager.py", line 274, in load_setuptools_entrypoints
    plugin = ep.load()
  File "c:\users\shhey\pyccd\.venv\lib\site-packages\pkg_resources\__init__.py", line 2410, in load
    self.require(*args, **kwargs)
  File "c:\users\shhey\pyccd\.venv\lib\site-packages\pkg_resources\__init__.py", line 2433, in require
    items = working_set.resolve(reqs, env, installer, extras=self.extras)
  File "c:\users\shhey\pyccd\.venv\lib\site-packages\pkg_resources\__init__.py", line 791, in resolve
    raise VersionConflict(dist, req).with_context(dependent_req)
pkg_resources.VersionConflict: (pytest 4.3.0 (c:\users\shhey\pyccd\.venv\lib\site-packages), Requirement.parse('pytest<4.0.0'))

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "c:\python36\Lib\runpy.py", line 193, in _run_module_as_main
    "__main__", mod_spec)
  File "c:\python36\Lib\runpy.py", line 85, in _run_code
    exec(code, run_globals)
  File "C:\Users\shhey\pyccd\.venv\Scripts\pytest.exe\__main__.py", line 9, in <module>
  File "c:\users\shhey\pyccd\.venv\lib\site-packages\_pytest\config\__init__.py", line 61, in main
    config = _prepareconfig(args, plugins)
  File "c:\users\shhey\pyccd\.venv\lib\site-packages\_pytest\config\__init__.py", line 196, in _prepareconfig
    pluginmanager=pluginmanager, args=args
  File "c:\users\shhey\pyccd\.venv\lib\site-packages\pluggy\hooks.py", line 289, in __call__
    return self._hookexec(self, self.get_hookimpls(), kwargs)
  File "c:\users\shhey\pyccd\.venv\lib\site-packages\pluggy\manager.py", line 68, in _hookexec
    return self._inner_hookexec(hook, methods, kwargs)
  File "c:\users\shhey\pyccd\.venv\lib\site-packages\pluggy\manager.py", line 62, in <lambda>
    firstresult=hook.spec.opts.get("firstresult") if hook.spec else False,
  File "c:\users\shhey\pyccd\.venv\lib\site-packages\pluggy\callers.py", line 203, in _multicall
    gen.send(outcome)
  File "c:\users\shhey\pyccd\.venv\lib\site-packages\_pytest\helpconfig.py", line 93, in pytest_cmdline_parse
    config = outcome.get_result()
  File "c:\users\shhey\pyccd\.venv\lib\site-packages\pluggy\callers.py", line 80, in get_result
    raise ex[1].with_traceback(ex[2])
  File "c:\users\shhey\pyccd\.venv\lib\site-packages\pluggy\callers.py", line 187, in _multicall
    res = hook_impl.function(*args)
  File "c:\users\shhey\pyccd\.venv\lib\site-packages\_pytest\config\__init__.py", line 655, in pytest_cmdline_parse
    self.parse(args)
  File "c:\users\shhey\pyccd\.venv\lib\site-packages\_pytest\config\__init__.py", line 871, in parse
    self._preparse(args, addopts=addopts)
  File "c:\users\shhey\pyccd\.venv\lib\site-packages\_pytest\config\__init__.py", line 817, in _preparse
    self.pluginmanager.load_setuptools_entrypoints("pytest11")
  File "c:\users\shhey\pyccd\.venv\lib\site-packages\pluggy\manager.py", line 280, in load_setuptools_entrypoints
    message="Plugin %r could not be loaded: %s!" % (ep.name, e),
pluggy.manager.PluginValidationError: Plugin 'profiling' could not be loaded: (pytest 4.3.0 (c:\users\shhey\pyccd\.venv\lib\site-packages), Requirement.parse('pytest<4.0.0'))!

Dear,
Is there introductory user guide for pyccd?
I'm wondering if pyccd can be used by other data than ARD? if it's the case could please provide the input data specification?

Thanks
Mustapha

Hi Kelcy Smith,

when i am running this code: https://github.com/USGS-EROS/lcmap-pyccd/blob/develop/notebooks/ccd_example.ipynb
I find i can't get the same result as you showed in this page. I am wondering why i can't get the same result even i use the default test data you used?

I still have one more question: how can i prepare my data for the CCD method? i am thinking get the scene list from Earthexplorer, and Download the dataset from ESPA. then input one pixel's time series data to this CCD method, and try to detect the break as you showed in this page: https://github.com/USGS-EROS/lcmap-pyccd/blob/develop/notebooks/ccd_example.ipynb

Thanks

It's int-casting count_clear_or_water(quality)/count_total(quality)

This means that a value like 0.99 evaluates to 0, IE. The unexpected result is that everything that doesn't have 100% clear coverage is now classified as insufficient coverage.

Quickest fix would be:

return np.float16(count_clear_or_water(quality))/np.float16(count_total(quality))

I'll make a PR shortly.

I've run into a rare case that throws a
numpy.linalg.linalg.LinAlgError: Singular matrix error.

To my understanding:

• The fit function uses a closed form regression technique to calculate the best fit.
• This regression technique relies on matrix inversion.
• Singular matrices do not have inverses.
• X' dot X might be singular.
• Hence this error is thrown.

My suggestion would be including a try catch that switches over to SGD if a closed form fails. I'll also need to revisit my data to see what sort of values are causing this error.

Dear,
Is there introductory user guide for pyccd?
I'm wondering if pyccd can be used by other data than ARD? if it's the case could please provide the input data specification?

Thanks
Mustapha

Dear,
when I followed the installation guidelines to install the device and test dependencies,it tell me

File "setup.py" not found. Directory cannot be installed in editable mode: /home/hanlele/pyccd
Could help me to solve this problem？
thank you very much!

Hey,

When i am using pyccd (https://github.com/USGS-EROS/lcmap-pyccd/tree/release/2017.06.20). I choose the test data, "test_3657_3610_observations.csv" in the resources folder. however, i got an error: ValueError: Unsupported bitpacked QA value 0

i am wondering why i got a error when i use the dataset you provided?

Thanks

I'm working on generating a "change-density" heat-map visualization using pyccd. Each pixel in this map is assigned a value equivalent to the number of change-models generated for that pixel using pyccd.

Here is a profile of my results:

• 15 years of landsat imagery (~220 acquisitions)
• 10km^2 extent (~300px ~300px)
• takes approximately 2.1 hrs on 8 cores
• single pyccd process takes .18 seconds (or 180 ms)

I was wondering what strategies USGS-EROS uses to run pyccd based analysis on larger areas?

ccd.detect doesn't seem to be working on your test data.

Here is the output that I'm getting:

{'algorithm': 'lcmap-pyccd:1.1.0', 'change_models': [], 'procedure': 'standard_procedure', 'processing_mask': array([False, True, False, True, True, False, False, True, False, True, True, True, True, True, False, False, False, True, False, True, False, False], dtype=bool)}

I would suggest reformatting the test data such that there is little chance that my "Formatting it to make it work in python" is the source of error.

What are the meaning and use of curve_qa values? I ran the code on test_3657_3610 and got a total of 5 models. The curve_qa value in the first four was 8 and in the last one was 24.