strax
strax copied to clipboard
AxFoundation
Stream analysis for xenon TPCs
## What is the problem? Data chunking is currently done semi-arbitrarily at two levels, the "run" and the "chunk". The concept of a run is not always relevant (e.g. MC...
We should make the the storage directory make a sub directory. For example: ``` # From 016389-raw_records-rfzvpzj4mf/ # To 016/016389-raw_records-rfzvpzj4mf/ ``` If we allow generic characters for the run-numbers, each...
We should add beside the already existing fuzzy_for option an option which allows to ignore Option/Versions changes for the whole dependency graph of a plugin. I just had the case...
We could add a function `st.take_snapshot` to store some seconds of data e.g. for paper plots of specific events. Could be exported as a new run or something that could...
Users should be aware that lines like the following: ``` for d in data_type: data= d['data'][:d['length']] modifies_data(data) ``` will create only a "view" onto data in memory. Hence when `data`...
**Please do not read** I still need to figure out if this is actually a problem or not. Nevertheless, let me write it down **Symptons** ```python # from straxen/tests/ import...
It would be nice to support loading data as dask.delayed dataframes (like hax.minitrees.load, or at least somewhat like it), so we can use its parallel and out of core computation...
Each plugin's mailbox times out processing when it has received no new data for too long, suspecting strax might be in deadlock. The amount of time per chunk can very...
Mailbox should be factored out into an independent repository rather than using something else. Closes #62.
Found by @WenzDaniel. If you call e.g. `st.make(run_id, 'bla', register=PluginMakingBla)`, you get a `KeyError: 'No plugin class registered that provides 'bla''`. This arises because `make` (unlike `get_array`) checks to see...