Punchout data representation

[alecandido]

We could get a more comprehensive representation of punchout data, just adding more plots. At the same time, this would lift the constraint on the current ones, such that we can always represent raw data - which are a more faithful representation of the experimental results.

Semi-Raw Data Plotting

In practice, pure raw data are hardly desired: if the data are hidden just behind a simple projection, then better to plot the suitably projected data.

For the punchout, we could apply this general idea for two instances:

• phase information: display the unwrapped phase with cable-delay subtracted
• (amplitude sweep-only) gradient: present signal after known gradient subtraction

Note

Why is there a gradient?

The reason essentially boils down to the usage of pulses.

How to subtract it?

Just divide by the amplitude.

If there is any point at zero amplitude, it can be safely discarded.

Filtered Data Visualization

Add plots that showcase filtered data as supplementary visualizations.

This could be useful to understand the fit behavior: since the peaks-finding algorithm just works on

[alecandido]

http://login.qrccluster.com:9000/rmi4JTQzSkKXzb96gQnlaA==/

A good reason not to represent only raw data:

We also want raw data, but the peaks are fitted on the data normalized row by row. And that would be useful to plot as well.