-
Notifications
You must be signed in to change notification settings - Fork 28
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Switch from Golang to Python #43
Conversation
…pod-autoscaler into switch_from_golang_to_python
Codecov Report
@@ Coverage Diff @@
## master #43 +/- ##
==========================================
- Coverage 75.79% 75.26% -0.53%
==========================================
Files 9 11 +2
Lines 252 283 +31
==========================================
+ Hits 191 213 +22
- Misses 60 65 +5
- Partials 1 5 +4
Flags with carried forward coverage won't be shown. Click here to find out more.
|
hey @jthomperoo! I just saw this, and I'm curious what other TODOs do you need to get this merged in? or is it just like final/edge case testing? thanks! |
Hey @kevjin - thanks for your other pull request btw, just having a look through now and I'll get it merged in - for this work to be honest I had left it as I didn't have much time. As far as I can remember it was 90% done, just wanted to test that it worked without breaking anything. Thanks for reminding me of this, I think I'll have a look at these changes at the weekend and see if I can get it merged in. Feel free to have a look through the changes. There was a request to add ARIMA forecasting to this (#37) which was on the to-do list after this Python stuff was added in, which would hopefully be easy enough with the Python libraries (easier than reimplementing in Golang anyway!). |
awesome thanks! I'm actually trying to get this working for my own LSTM model haha, so I'll definitely be looking, and doing some of my own manual testing on this PR. I'll let you know if I find any bugs! |
algorithm/algorithm.go
Outdated
|
||
const ( | ||
entrypoint = "python" | ||
shellTimeout = 10000 |
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.
imo a hard 10 second timeout is a little too restrictive. Some models, like ARIMA, will probably take longer to run. Also the more storedValues
you have, the longer it'll take the model to run. What about 30 seconds?
Although, I'm not sure if the CPA would respect the longer timeout (haven't really looked at it)
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.
Yep good point, I've made a bunch of changes so that this value is configurable, and upped the default timeout to 30s.
Resolves #38