disperse_output_reader

Simple functions to read Disperse output and save it in hdf5

Basic usage

Import this file as python module (or simply copy-paste functions in your code).

Reading

To read in a Disperse NDSkl_ascii file:

read_skeleton_from_NDskl_ascii(filename, as_dataframes=False)

where filename (a string) is the name of (plus full path to) the Disperse file.

This will return a class containing all the data as fields, as shown below. If as_dataframes=True, this structure will contain two additional members (df_critical_points and df_filaments) containing the same information (except the int values) but organized in a pandas dataframe

[for critical points]

• Ndim : [int] number of dimensions
• bbox : [list of list of float] bounding box of the data, format: [[x0, y0, z0], [x1, y1, z1]] where 0 and 1 indicate two opposite corners
• Ncritical_points : [int] number of critical points
• critical_points : [list of np.array] coordinates of the critical points, format: [..., np.array([x_i, y_i, z_i]), ...], where i indicates the i-th critical point
• cp_type : [list of int] crtical point type
• cp_value : [list of int] value of the DTFE-computed density at the critical point
• cp_pairID : [list of int] persistence pair index for each critical point
• cp_boundary : [list of int] boundary flag for each critical point
• Nconnected_fil : [list of int] number of filaments connected to each critical point
• other_cp_id : [list of list of int] for each critical point, a list of the critical point at the opposite end of each connected filament
• filament_id : [list of list of int] for each critical point, a list of the indices of the associated filaments, indexing the arrays in the filament section (below)
• Nfields_cp : [int] number of additional fields associated to each critical point
• fields_cp_name : [list of string] name of the additional fields associated to each critical point
• fields_cp_values : [np.array with shape (Ncritical_points, Nfields_cp)] values of the additional fields associated to each critical point

[for filaments]

• Nfilaments : [int] number of filaments
• Nsamples : [list of int] number of sampling points for each filament
• filaments : [list of list of Ndim float] for each filament, a list of coordinates of the sampling points
• cp_extremes : [list of int pairs] for each filament, the indeices of the two extrema, indexes the critical point fields
• Nfields_fil : [int] number of additional fields associated to each filament
• fields_fil_name : [list of string] name of the additional fields associated to each filament
• fields_fil_values : [np.array with shape (Ncritical_points, Nfields_cp)] values of the additional fields associated to each filament

Writing

To write in an HDF5 file the content of the class described above, use:

save_skeleton_to_hdf5(skeleton, filename_hdf5)

where skeleton is the class to write and filename_hdf5 is the complete name of the HDF5 file to create.

The format for this file is described at: https://thesan-project.com/filaments.html