Fixes for eventbrowser #557
Conversation
…rnog_eventbrowser Conflicts: changelog.txt
…dioMC into rnog_eventbrowser
…dio/NuRadioMC into rnog_eventbrowser
…ioMC into RNOG_data_example
…dio/NuRadioMC into RNOG_data_example
…o RNOG_data_example
…RNOG_data_example
…o RNOG_data_example
…line data browser?)
…o rnog_eventbrowser
|
@fschlueter I/we may need to have another look at how to merge this with #583, I think I prefer my solution for the race condition on the file handler, but there are a couple of things in your implementation that are also nice to have. I don't think I'll manage to do this today, though, so my suggestion would be to merge this first |
|
@sjoerd-bouma, yes your solution looks cleaner. I haven't checked if it works similar to my solution did you? How many times is the io class spawned when you read a single file (single user)? What happens if the user_id changes from None to a hash? We should also merge the 3 dataprovider files/classes into one! Can you do that? Should we sit together when I am in Erlangen? |
|
@sjoerd-bouma thanks for doing that! I noticed that the data handler spawns still two instances of the io class (once for user None and once for the user with a specific hash). Do you think we could fix that or not? Otherwise this looks good to me! |
|
@fschlueter Yeah, I guess that would be good to fix, but I'm not sure how straightforward that would be... the user_id is set by a callback, and they fire in parallel threads with no way of forcing that one to fire before all the others, I think. |
|
Okay, you could think about copying the object rather than initialising a new one once a none None user_id is passed. But I am fine with how it is right if you want to merge it. |
|
Sorry @fschlueter, I created a merge conflict so you'll have to approve again. I also hopefully 'fixed' the user_id=None issue - if there is an entry for user None present this gets 'recycled' by the first user with an actual id. |
refactor few lines
| " of mattak (https://github.com/RNO-G/mattak)." | ||
| ) | ||
| reader = _readRNOGData_eventbrowser(load_run_table=False) | ||
| reader.begin([os.path.dirname(f) for f in filename], overwrite_sampling_rate=3.2) |
There was a problem hiding this comment.
By now the mattak reader supports singles files. It's somewhat a dirty hack but should work? Furthermore I have a PR in mattak open which would implement "native" support for singles files.
This branch contains a couple of fixes for the eventbrowser:
This last feature also requires an update to dash>=2.9.2
The branch is pretty old and seems to include more 'features' than I realized (e.g. an old RNO-G data example), so I probably have to clean this up before we can merge this.