Patch progress bar

[WenzDaniel]

With this pull request we add a nice progessbar. Implemented in get_iter. It also works if the user specifies any time range.

[JoranAngevaare]

Hi Daniel, would it be an option to make the progress-bar optional? Like e.g. https://github.com/JelleAalbers/wimprates/blob/master/wimprates/utils.py#L44.

[WenzDaniel]

Hi Joran, thanks. I have to check.

[WenzDaniel]

Hi Daniel, would it be an option to make the progress-bar optional? Like e.g. https://github.com/JelleAalbers/wimprates/blob/master/wimprates/utils.py#L44.

Done, the bar is now optional. By default the bar is switched on. Also added a check that the progress bar is only used if metadata exists.

[JoranAngevaare]

@WenzDaniel some extra questions:

• Do you know how much extra time we need for a typical run to load the data (and does it differ much for event_info / raw_records)?
• If so perhaps we can make the default different for get_array/get_df might be an option.
• Do you know if it also works for superruns or should we always force it to be false there?

[WenzDaniel]

@WenzDaniel some extra questions:

* Do you know how much extra time we need for a typical run to load the data (and does it differ much for event_info / raw_records)?

To be honest I never really checked, since the overall added extra time is super small compared to all the other checks we are doing when reading data. I expect the time difference to be very small. I guess the part which takes longest is the reading in of the metadata for which I have not seen any big differences for the different datakinds.

Update:

I just checked with

%%timeit
for i in st.get_iter(run_id_led, 'raw_records', seconds_range=(0, 30), progress_bar=True):
    pass

for a typical calibration run and I got the following values:

• 1min 44s ± 1.33 s
• 1min 47s ± 2.76 s

So there seems to be an overhead O(1 s) which I think is fine.

* Do you know if it also works for superruns or should we always force it to be false there?

No, I never worked with superruns and I do not know how they work. If it is a nested loop for run in runid -> for .. in get_iter -> you will get for each run an individual progress bar, if you first somewhat make a "super run" for which you loop just once with get_iter you will get just a single bar.

** Update: **

I just read through https://github.com/AxFoundation/strax/blob/master/strax/run_selection.py#L221 and since super runs have the same structure as regular runs including a metadata file the progess bar should also work for them.

[JoranAngevaare]

@ ACs this is ready for merge unless @WenzDaniel feels like addressing CodeFactor

[JoranAngevaare]

@WenzDaniel , this PR would be great but I cannot circumvent the codefactor