Conversation
|
Check out this pull request on See visual diffs & provide feedback on Jupyter Notebooks. Powered by ReviewNB |
znicholls
left a comment
znicholls
left a comment
There was a problem hiding this comment.
Looks a good start. I guess it makes sense to get this in, then switch to some sort of xarray backend (dealing with #127 in the process) and then we'd have a pretty clear path to dealing with some sort of lazy loading over lots of runs..?
| @@ -0,0 +1,226 @@ | |||
| """ | |||
| :class:`ScmEnsemble` provides a container for holding data from different model runs. | |||
There was a problem hiding this comment.
| :class:`ScmEnsemble` provides a container for holding data from different model runs. | |
| :class:`ScmEnsemble` provides a container for holding data from different model runs. | |
| This is particularly useful for holding a collection of data on different time axes, without | |
| having to put all the data on the same time axes first (which can be a moderately expensive | |
| operation). |
| """ | ||
| str: Name of meta containing a unique identifer for each run | ||
|
|
||
| If no value is provided, then an implicit value is create. This ensures that every timeseries |
There was a problem hiding this comment.
| If no value is provided, then an implicit value is create. This ensures that every timeseries | |
| If no value is provided, then an implicit value is created. This ensures that every timeseries (across all :obj:`ScmRun` objects) |
| items.append(r[item]) | ||
| except KeyError: | ||
| continue | ||
| if not len(items): |
There was a problem hiding this comment.
| if not len(items): | |
| if not items: |
| pdt.assert_frame_equal( | ||
| ensemble.runs[0].timeseries().reset_index(), | ||
| res.xs(0, level="run_id").dropna(how="all", axis="columns").reset_index(), | ||
| check_like=True, # To remove once #97 is resolved |
There was a problem hiding this comment.
| check_like=True, # To remove once #97 is resolved | |
| check_like=True, |
What is #97? Can be removed right?
|
|
||
| assert res.run_ids == [0, 1] | ||
|
|
||
| # TODO: fix inconcistency with columns type of timeseries |
|
|
||
| def __init__(self, runs=None): | ||
| """ | ||
|
|
There was a problem hiding this comment.
Are we intending on adding more loading functionality? If no, we should change the notebook
There was a problem hiding this comment.
Just looked at the notebook properly, assuming we're waiting to sort this before we start mucking around with writing the notebook?
2 participants
Pull request
Please confirm that this pull request has done the following:
CHANGELOG.rstaddedThoughts before I flesh out docs
Closes #98