Maziyar Boustani created CLIMATE-322:
----------------------------------------
Summary: dataset_processor._get_subregion_slice_indices cannot
handle imprecise start and end time input
Key: CLIMATE-322
URL: https://issues.apache.org/jira/browse/CLIMATE-322
Project: Apache Open Climate Workbench
Issue Type: Improvement
Components: regridding/data processing
Affects Versions: 0.3-incubating
Reporter: Maziyar Boustani
Assignee: Maziyar Boustani
Fix For: 0.4-incubating
Similar to (https://issues.apache.org/jira/browse/CLIMATE-321), but this issue
is about not finding start and end time from dataset date list.
Error Massage:
File
"/Users/MBoustani/Documents/code/python/program/RCMES_cmd/trunk/ocw/dataset_processor.py",
line 708, in _get_subregion_slice_indices
timeStart = np.nonzero(target_dataset.times == subregion.start)[0][0]
IndexError: index 0 is out of bounds for axis 0 with size 0
Problem:
============================
The problem is that code is looking for exact time given from user into dataset
time list. If the exact time is not matched (found) then we get the error.
Solution:
============================
It is better to find the closest time from available dataset time from user
given time.
The code will always get closest time after given start time and closest time
before given end time to always have safe time selection.
Example:
============================
Actual dataset start time: 2000-01-01
Actual dataset end time: 2005-12-01
If user set start time as 2000-01-03, then the current code will break due to
not finding the exact 2000-01-03 from dataset date list.
Therefore code should find the closest date after user given start time
(2000-02-01).
Same thing with end time, if the user set end time as 2005-11-20, then code
should pick closest date before end time (2005-11-01).
--
This message was sent by Atlassian JIRA
(v6.1#6144)
