[
https://issues.apache.org/jira/browse/CLIMATE-643?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Huikyo Lee updated CLIMATE-643:
-------------------------------
Description:
Paul L. suggested some ideas to update examples. For example,
"knmi_to_cru_full_bias.py" needs to be updated with better description. The
model to model bias could be replaced by model to observation data bias. The
goal is providing 5 examples all based on an actual published papers.
Currently, OCW examples generate wrong results when there is missing data in
observational datasets. It is important to mask those grid points with missing
values in model datasets so that no metrics calculation is done at those grid
points. In other words, if any of observation/model dataset has missing value
at a grid point, we need to propagate the missing information to the other
datasets.
was:
Paul L. suggested some ideas to update examples. For example,
"knmi_to_cru_full_bias.py" needs to be updated with better description. The
model to model bias could be replaced by model to observation data bias. The
goal is providing 5 examples all based on an actual published papers.
Currently, OCW examples generate wrong results when there is missing data in
observational datasets. It is important to mask those grid points with missing
values in model datasets so that no metrics calculation is done at those grid
points. In other words, if any of observation/model dataset has missing value
at a grid point, non-missing values in the other datasets need to be masked.
> Updating some of examples
> -------------------------
>
> Key: CLIMATE-643
> URL: https://issues.apache.org/jira/browse/CLIMATE-643
> Project: Apache Open Climate Workbench
> Issue Type: Improvement
> Components: general
> Affects Versions: 1.0.0
> Reporter: Huikyo Lee
> Assignee: Huikyo Lee
> Fix For: 1.0.0
>
>
> Paul L. suggested some ideas to update examples. For example,
> "knmi_to_cru_full_bias.py" needs to be updated with better description. The
> model to model bias could be replaced by model to observation data bias. The
> goal is providing 5 examples all based on an actual published papers.
> Currently, OCW examples generate wrong results when there is missing data in
> observational datasets. It is important to mask those grid points with
> missing values in model datasets so that no metrics calculation is done at
> those grid points. In other words, if any of observation/model dataset has
> missing value at a grid point, we need to propagate the missing information
> to the other datasets.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
