Debug for EventwBayesClass because peaks overlapping

[dachengx]

Before you submit this PR: make sure to put all operations-related information in a wiki-note, a PR should be about code and is publicly accessible

What does the code in this PR do / what does it improve?

Debug EventwBayesClass. In very rare cases the peaks' "time" can be the same, so simply use the matching between time

straxen/straxen/plugins/events/event_w_bayes_class.py

Lines 29 to 30 in c618480

	mask = np.in1d(peaks["time"], events[f"{name}_time"])
	mask_ev = np.in1d(events[f"{name}_time"], peaks["time"])

can cause error like:

CRITICAL:ThreadedMailboxProcessor:Target Mailbox (event_w_bayes_class_paired) killed, exception <class 'strax.mailbox.MailboxKilled'>, message (<class 'ValueError'>, ValueError('NumPy boolean array indexing assignment cannot assign 48003 input values to the 47986 output values where the mask is true'), <traceback object at 0x7f62537a4e00>)

Can you briefly describe how it works?

Mimicking what EventAmbience does, split the peaks for each event and assign the fields based on indices:

straxen/straxen/plugins/events/event_ambience.py

Lines 42 to 57 in c618480

	def compute(self, events, peaks):
	split_peaks = strax.split_by_containment(peaks, events)
	# 1. Initialization, ambience is set to be the lowest possible value
	result = np.zeros(len(events), self.dtype)
	# 2. Assign peaks features to main S1, main S2 in the event
	for event_i, (event, sp) in enumerate(zip(events, split_peaks)):
	for idx, main_peak in zip([event["s1_index"], event["s2_index"]], ["s1_", "s2_"]):
	if idx >= 0:
	for ambience in self.origin_dtype:
	result[f"{main_peak}n_{ambience}"][event_i] = sp[f"n_{ambience}"][idx]
	# 3. Set time and endtime for events
	result["time"] = events["time"]
	result["endtime"] = strax.endtime(events)
	return result

.

The results are not changed so I will not update the __version__.

Can you give a minimal working example (or illustrate with a figure)?

Please include the following if applicable:

• Update the docstring(s)
• Update the documentation
• Tests to check the (new) code is working as desired.
• Does it solve one of the open issues on github?

Notes on testing

• Until the automated tests pass, please mark the PR as a draft.
• On the XENONnT fork we test with database access, on private forks there is no database access for security considerations.

All italic comments can be removed from this template.

[dachengx]

However, this plugin is not widely used.

[WenzDaniel]

I am wondering whether we should not remove this plugin. Is it still being maintained? I do not think so.

[dachengx]

I am wondering whether we should not remove this plugin. Is it still being maintained? I do not think so.

For the reproducibility of old results, we still need it. I think we can keep it for a while.

[WenzDaniel]

Old results use old container no? Otherwise it is not reproducible anyhow?

[dachengx]

Old results use old container no? Otherwise it is not reproducible anyhow?

Yes. You are right. I mean the reproducibility of the codes in other repo which still uses this plugin.

[coveralls]

Changes unknown
when pulling 166c64c on debug_event_w_bayes_class
into ** on master**.

[MerzJohannes]

Looks good to me. If we want to keep it for reproducibility we can merge it.

[yuema137]

In the previous code, mask may have a larger length than mask_ev (and events) when there are duplicated 'time' in peaks. The PR fixes this problem by looping over split peaks together with events, and only assigning the probability when a valid peak index is found. It looks good to me.