Parsing point name in nd2

[hftsai]

Hi Great tools you have created here
I'm a python novice, but I'm currently thinking to automate my experiment process.

We are conducting high throughput experiments with nikon microscopes
one way to annotate experiments during a ND acqusitions is to rename the point name (as screen shot attached)

Is there a function in nd2reader to parse the point name in a multipoint loop?

Thanks!
Best,

[rbnvrw]

Thanks @hftsai, I'm not sure if these point names are parsed at this moment, but you can have a look in the attributes in reader.parser._raw_metadata (where reader is your ND2Reader instance). For example maybe the attributes custom_data, image_attributes, image_metadata_sequence or image_metadata may contain the information (no guarantees though).

Let me know if you find something!

[hftsai]

Thank you!
it doesn't seem to be in the raw_metadata or in the metadata.

I'll look into the Nikon SDK

Thanks!

[aknckaan]

I was having the same issue then I noticed that the point names are saved at reader.parser_raw_metadata.image_metadata['SLxExperiment']["uLoopPars"]["Points"]

[rbnvrw]

Great that you found this. The discoverability of all these attributes is quite low at this point. It is basically an almost one-to-one mapping of the ND2 file format into the raw metadata class. It would be great if we can improve this in future versions. I'll leave this open as a reminder.

[kevinjohncutler]

I would find this enhancement extremely useful - filtering data by point name is something I do regularly. Until then, @aknckaan's comment is workable. One thing I had to contend with is that reader.parser._raw_metadata.image_metadata gives byte objects, not strings, as the keys and values. One can recursively convert or just deal with typing b'Points' etc., but unless I'm just being stupid, an accessible version of this data would be in str form.