Faulty X Axis Scaling in Qviewkit

[fr34q]

When using logarithmical spaced values as x axis in a measurement (e.g. for a T1 measurement) the viewer interprets the x axis as lineary spaced thus leading to wrong axis labelling.

Example: data0 > delay: (x axis is not really the delay but index) This leads to a T1 measurement looking like: data0 > phase_avg_0 instead of an exponential decay

Expected behavior: When x axis is labelled "delay" it should also use the values given in the delay dataset and not extrapolate from only the first two data points or so.

[rotzinger]

Hi Richard,

this is most likely not a bug. Datasets in data0/analysis0 are displayed in Qviewkit as bare dataset with the x-axis being either None (just the index) or a optional x0+dx*N pair. If you want to display a dataset vs a different axis use the add_view() function in qkit.storage.store , which creates a view on the data.

Cheers, Hannes

[fr34q]

Hi Hannes,

shouldn't we then at least update the x axis labelling accordingly to avoid confusion? In my case, I expected to see an exponential decay and it took me a while to realize what was happening here because the x axis label was given as "delay (ns)".

But as we typically set the x axis explicitly I would find it more intuitive to use if the hdf data container would then contain these views automatically. Alternatively, a nice solution would be if one could specify an axis dataset for each individual dataset which is then used instead of linear spacing.

Your thoughts?

See you, Richard

[Schneider1]

Although considered to be not a bug, this behavior should have changed. Can you confirm that, Richard?

[fr34q]

Yes, as I said last week, this now works as I would expect it.

Delays for T1 measurement:

Plot of the averaged phase:

So issue can stay closed ;)