Merged changes to peaks/dev into peaks_AndersDev, and added the sgm4 …

[AndersSMortensen]

Merging the sgm4 loader into the peaks dev branch.

[shum23]

Thanks @AndersSMortensen for the contribution and changes and sending some test data.

Note: the following has been fixed in commit c2188e8. Putting discussions here for future reference.

• The timestamp seems to exist in the data files under the key 'Entry/Data/Timestamp', but the timestamp that shows in the metadata is the file creation timestamp on the system, which is the fallback in peaks​ when no timestamp is available in the data. It would be great to map the initial timestamp from the data (i.e. the start time) to the metadata.scan.timestamp​ field
• metadata.scan.loc​ is SGM4 now. To be consistent with the general naming conventions used elsewhere in peaks​, would you be able to rename it to e.g. ASTRID2_SGM4
• analyser.scan.eV​ always calculates the centre kinetic energy. This is great for fixed mode scans, but for swept scans, it could be helpful to show the energy range as a two-value array [E_start, E_end] as is implemented in other peaks​ loaders?
• You have a scan with a 'Loop' co-ordinate. If we understand correctly, that is basically a repetition of the same scan just at e.g. a later point in time, making multiple iterations. This behaviour is already encoded elsewhere in peaks​ using a dimension name of scan_no ​(see, e.g. _desired_dim_order​ of class:BaseARPESDataLoader​, as e.g. used in the SES data loader. Given that this seems really the same functionality here, it would be helpful if we could maintain the scan_no​ co-ordinate/dim name for this too, if possible (I think should just be renaming Loop to scan_no)?
• We had some complications over the sign conventions for loaders before. So can we just double check, for 'WSe2Cold_4_0' or 'WSe2Cold_7_0', should the actual normal emission be approximately:
  manipulator polar: -26 deg
  tilt: -9 deg
  azi: 10 deg if aligned with gamma-M
  The key question there is whether the signs for those are as you would expect for this data set or are inverted.
• A trivial thing, but there are a couple of lines commented out in the loader file, I guess used for debugging or inherited from the diamond loader. If you have a chance, might be good to clean them.

[shum23]

Hi @AndersSMortensen,

Thanks. I think this looks great and is ready to merge. I've just noticed a couple of wee typos about the Keithley parameters. Around line 633 in sgm4.py, KA_V seems to map to KeithleyA_Resistance while on line 116 it's pulling the keithleyA_Voltage metadata. I reckon it means the voltage there, and the same for KA_R, KB_V, and KB_R. If you could fix these typos that'd be great.

And would you also be able to add an entry for this new loader in the changelog?

[AndersSMortensen]

Thank you @shum23! I have fixed the typos and added a comment in the changelog.
I have just written that the loader is there. If there is any key information you would like me to add, let me know.

[shum23]

Thanks @AndersSMortensen. Merging now!