CSV

The pynxtools-apm parser and normalizer reads the following content and maps them on respective NeXus concepts that are defined in the NXapm application definition:

CSV	NeXus/HDF5
Reconstructed positions (x, y, z)
Mass-to-charge-state-ratio values (m/q)

Occasionally, atom probers export portions of a reconstruction as comma-separated ASCII files. We consider this to be an unnecessary inefficient and inaccurate approach that should be replaced by self-descriptive formats like NeXus. In every case, at least a header should be specified which details the information that is stored in each column but examples from the literature show that this is not always supplied.

The ifes_apt_tc_data_modeling library makes strong assumptions when trying to read atom probe content from a CSV file: - The format is comma-separated text - Splitting on the separator results in a table with one row for each ion and four columns. - The first three columns in sequence are mapped to reconstructed positions (x, y, z) i.e. like in POS - The fourth column is mapped to mass-to-charge-state-ratio values

Users of this parsing functionality should inspect carefully whether the results are to their expectation. If this is not the case, feel free to submit an issue issue here so that we can inspect and eventually fix this issue or guide how to use possible alternatives.