-
Notifications
You must be signed in to change notification settings - Fork 15
Conversation
@mzuehlke You're the expert on observing monitors! Can you please check this out? I felt a bit unsure in places if what I'm doing is the correct thing to do. Especially in correlation (monitor.observing encompasses a lot of code) and in index.py and timeseries.py. |
I will take a look |
Codecov Report
@@ Coverage Diff @@
## master #430 +/- ##
==========================================
+ Coverage 75.16% 75.22% +0.06%
==========================================
Files 72 72
Lines 10268 10293 +25
==========================================
+ Hits 7718 7743 +25
Misses 2550 2550
Continue to review full report at Codecov.
|
I only looked at the operators you had modified and improved the code in some places.
|
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Feel free to merge
cate/ops/correlation.py
Outdated
r.values[r.values < -1.0] = -1.0 | ||
r.values[r.values > 1.0] = 1.0 | ||
with monitor.child(1).observing("task 1"): | ||
negativ_r = r.values < -1.0 |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@mzuehlke
So, we only monitor when we try to access the values? As in, processing from lines 239-247 would be deferred until then?
EDIT: Oh my god I'm dumb, I just read my own comment.
I will in a calm moment try to make the dask progress monitor a bit smarter. Because I find this multiple observing calls not the nicest solution..... |
So, for future reference - as it is now, we can only wrap a single line in |
Currently: yes |
What the title says.