[feature] : Handle force curve data in a unified manner for all formats?

[SylviaWhittle]

Is your feature request related to a problem?

Curve data is stored in several file formats, would be good to be able to access and unpack this.

Describe the solution you would like.

A unified way (less jank per implementation) of retrieving the curve data from each loader function. @ahobbs7 is working on this for .jpk-qi force curves and thought maybe we want to generalise it for other potential implementations.

Discuss here maybe?

Describe the alternatives you have considered.

No response

Sample Image

No response

Additional context

No response

[ahobbs7]

The current approach I have drafted is having a three dataclasses.

1. CurvesDataset is the object that is returned from any load_x() function where the file contains curve data. It contains a dictionary of CurvesVolume objects. Many files containing curve data that are read by AFMReader such as jpk data will just contain a single CurvesVolume named Trace. However, it seems some files such as ardf data contains separate trace and retrace curve volumes. Also, it allows for curves to be processed into a corrected format, be saved with the 'raw' data to a HDF5 format and be read from the same Dataset object with consistent metadata, but separated by being in a different volume. The other option here might be just treating corrected data as different channels (e.g. vDeflection and vDeflection_corrected) which could reduce the need for storing different volumes in a dictionary. It also contains a metadata attribute that is a CurvesMetadata instance for the data.
2. CurvesVolume is where the curves for different channels are accessible. It has both a get_curve(y, x) and a custom getitem method that allows for access like a numpy array e.g.

curve = volume[43, 54]

3. CurvesMetadata is where all the metdata is accessible including global metadata (the same across all curves or the settings of the AFM) and the metadata for both individual curves and individual segments of curves (i.e. approach, retract).

global_metadata = metadata.toplevel
curve_metadata = metadata[43, 54] # metadata for curve (43, 54) irrespective of segment
segment_metadata = metadata[43, 54, 0] # metadata for specifically the approach of curve (43, 54)

For each data format, child classes could be created to overide the get_curve or get_point_metadata function so that they retrieve the data from the underlying file format. CurvesDataset may not need a child class for each format depending on the case