lex issueshttps://code.ornl.gov/LIVVkit/lex/-/issues2019-05-02T18:54:02Zhttps://code.ornl.gov/LIVVkit/lex/-/issues/4Turn LEX analyses into a class2019-05-02T18:54:02ZKennedy, Joseph Hkennedyjh@ornl.govTurn LEX analyses into a classThis would allow us to define some default functions(methods) that users could choose to over-ride.
For example, we could have a default `parse_config` function that looks like:
```python
def parse_config(config: dict) -> argparse.Na...This would allow us to define some default functions(methods) that users could choose to over-ride.
For example, we could have a default `parse_config` function that looks like:
```python
def parse_config(config: dict) -> argparse.Namespace:
"""Generate an arguments Namespace from the analysis config file"""
test_args = test_args = OrderedDict([(k.replace('-', '_'), v) for k, v in config.items()])
test_args = argparse.Namespace(**test_args)
return test_args
```
Which spoofs the user defining a `parse_args` function and parsing command-line options with argparse.https://code.ornl.gov/LIVVkit/lex/-/issues/3Handle extension dependencies2019-02-12T16:39:12ZKennedy, Joseph Hkennedyjh@ornl.govHandle extension dependenciesBecause each extension is likely to introduce new dependencies, there needs to be a uniform method to handle the dependencies.
**Importantly, it's not feasible to create a LEX conda package because of the size of the data, unless, the ...Because each extension is likely to introduce new dependencies, there needs to be a uniform method to handle the dependencies.
**Importantly, it's not feasible to create a LEX conda package because of the size of the data, unless, the code and data were handled independently.**
Options include:
## Conda meta-package
By creating a conda meta-package all the dependencies could be handled via conda and this repo can be installed independently.
## Simply list in a `conda.yml` file
This is simpler than creating a meta-package, but requires people to manage their environments directly -- something that doesn't work well for ESMs as their analysis environments are usually set up by a sys. admin, and would be hard to include LEX dependencies.
## Adopt a data-only storage idea for this repository
Move the analysis code into LIVVkit proper, leaving only the data files behind. This will, however, make LIVVkit a much bigger package than needed for EVV and simple verification use.Kennedy, Joseph Hkennedyjh@ornl.govKennedy, Joseph Hkennedyjh@ornl.govhttps://code.ornl.gov/LIVVkit/lex/-/issues/2E3SM analyses should use ice sheet area2018-11-17T01:07:33ZKennedy, Joseph Hkennedyjh@ornl.govE3SM analyses should use ice sheet areaCurrently the E3SM analyses use the land gridcell area to weight the calculations, but really should be using the fractional ice-sheet area. This fractional area is usually contained in the `gris_area` or `aais_area`, but in the current ...Currently the E3SM analyses use the land gridcell area to weight the calculations, but really should be using the fractional ice-sheet area. This fractional area is usually contained in the `gris_area` or `aais_area`, but in the current E3SM compsets, these are all zeroed out.v0.2.0Kennedy, Joseph Hkennedyjh@ornl.govKennedy, Joseph Hkennedyjh@ornl.govhttps://code.ornl.gov/LIVVkit/lex/-/issues/1Added "_" in README2018-07-02T14:41:57ZEvans, KatherineAdded "_" in READMEI added text in README to develop, but forgot "_" at end of link, but it will not let me push to repo suddenly, so fix typo at end of link to README and look over text to see if its sufficient.I added text in README to develop, but forgot "_" at end of link, but it will not let me push to repo suddenly, so fix typo at end of link to README and look over text to see if its sufficient.Kennedy, Joseph Hkennedyjh@ornl.govKennedy, Joseph Hkennedyjh@ornl.gov