track_analysis

Track finding and fitting functions are listed here.

Methods

testbeam_analysis.track_analysis.find_tracks(input_tracklets_file, input_alignment_file, output_track_candidates_file, min_cluster_distance=False, chunk_size=1000000)

Takes first DUT track hit and tries to find matching hits in subsequent DUTs. The output is the same array with resorted hits into tracks. A track quality is set to be able to cut on good (less scattered) tracks. This function is uses numba to increase the speed on the inner loop (_find_tracks_loop()).

This function can also be called on TrackCandidates arrays. That is usefull if an additional alignment step was done and the track finding has to be repeated.

Parameters:

• input_tracklets_file (string) – Input file name with merged cluster hit table from all DUTs (tracklets file) Or track candidates file.
• input_alignment_file (string) – File containing the alignment information
• output_track_candidates_file (string) – Output file name for track candidate array
• min_cluster_distance (iterable, boolean) – A minimum distance all track cluster have to be apart, otherwise the complete event is flagged to have merged tracks (n_tracks = -1). This is needed to get a correct efficiency number, since assigning the same cluster to several tracks is not implemented and error prone. If it is true the std setting of 200 um is used. Otherwise a distance in um for each DUT has to be given. e.g.: For two devices: min_cluster_distance = (50, 250) If false the cluster distance is not considered. The events where any plane does have hits < min_cluster_distance is flagged with n_tracks = -1
• chunk_size (uint) – Chunk size of the data when reading from file.

testbeam_analysis.track_analysis.fit_tracks(input_track_candidates_file, input_alignment_file, output_tracks_file, fit_duts=None, selection_hit_duts=None, selection_fit_duts=None, exclude_dut_hit=True, selection_track_quality=1, pixel_size=None, n_pixels=None, beam_energy=None, material_budget=None, add_scattering_plane=False, max_tracks=None, force_prealignment=False, use_correlated=False, min_track_distance=False, keep_data=False, method='Fit', full_track_info=False, chunk_size=1000000)

Fits either a line through selected DUT hits for selected DUTs (method=Fit) or uses a Kalman Filter to build tracks (method=Kalman). The selection criterion for the track candidates to fit is the track quality and the maximum number of hits per event. The fit is done for specified DUTs only (fit_duts). This DUT is then not included in the fit (include_duts). Bad DUTs can be always ignored in the fit (ignore_duts).

Parameters:

• input_track_candidates_file (string) – Filename of the input track candidate file.
• input_alignment_file (string) – Filename of the input alignment file.
• output_tracks_file (string) – Filename of the output tracks file.
• fit_duts (iterable) – Specify DUTs for which tracks will be fitted. A track table will be generated for each fit DUT. If None, all existing DUTs are used.
• selection_hit_duts (iterable, or iterable of iterable) – The duts that are required to have a hit with the given track quality. Otherwise the track is omitted. If None: require all DUTs to have a hit, but if exclude_dut_hit = True do not use actual fit_dut. If iterable: use selection for all devices, e.g.: Require hit in DUT 0, and 3: selection_hit_duts = (0, 3). If iterable of iterable: define dut with hits for all devices seperately, e.g. for 3 devices: selection_hit_duts = ((1, 2), (0, 1, 2), (0, 1))
• selection_fit_duts (iterable, or iterable of iterable or None) – If None, selection_hit_duts are used for fitting. Cannot define DUTs that are not in selection_hit_duts, e.g. require hits in DUT0, DUT1, DUT3, DUT4 but do not use DUT3 in the fit: selection_hit_duts = (0, 1, 3, 4) selection_fit_duts = (0, 1, 4)
• exclude_dut_hit (bool) – Set to not require a hit in the actual fit DUT (e.g.: for unconstrained residuals). False: Just use all devices as specified in selection_hit_duts. True: Do not take the DUT hit for track selection / fitting, even if specified in selection_hit_duts.
• max_tracks (uint) – Take only events with tracks <= max_tracks. If None, take any event.
• force_prealignment (bool) – If True, use pre-alignment, even if alignment data is availale.
• selection_track_quality (uint, iterable) – One number valid for all DUTs or an iterable with a number for each DUT. 0: All tracks with hits in DUT and references are taken 1: The track hits in DUT and reference are within 2-sigma of the correlation 2: The track hits in DUT and reference are within 1-sigma of the correlation Track quality is saved for each DUT as boolean in binary representation. 8-bit integer for each ‘quality stage’, one digit per DUT. E.g. 0000 0101 assigns hits in DUT0 and DUT2 to the corresponding track quality.
• pixel_size (iterable of tuples) – One tuple per DUT describing the pixel dimension (column/row), e.g. for two FE-I4 DUTs [(250, 50), (250, 50)]. Only needed for Kalman Filter.
• n_pixels (iterable of tuples) – One tuple per DUT describing the number of pixels in column, row direction e.g. for 2 DUTs: n_pixels = [(80, 336), (80, 336)]. Only needed for Kalman Filter.
• beam_energy (uint) – Energy of electron beam in MeV. Only needed for Kalman Filter.
• material_budget (iterable) – Material budget of all DUTs. The material budget is defined as the thickness (sensor + other scattering materials) devided by the radiation length (Silicon: 93700 um, M26(50 um Si + 50 um Kapton): 125390 um). Only needed for Kalman Filter.
• add_scattering_plane (dict) –

  Specifies additional scattering planes in case of additional DUTs which are not used. The dictionary must contain the following keys:

  z_scatter: z position of scattering plane in um material_budget_scatter: material budget of scattering plane alignment_scatter: list which contains alpha, beta and gamma angles of scattering plane.

  If None, no rotation will be considered.

  In case of multiple scattering planes, each value of a key is a list, with items corresponding to each scattering plane. If add_scattering_plane is False, no scattering plane will be added.

• use_correlated (bool) – Use only events that are correlated. Can (at the moment) be applied only if function uses corrected Tracklets file.
• keep_data (bool) – Keep all track candidates in data and add track info only to fitted tracks. Needed for purity calculations!
• method (string) – Available methods are ‘Kalman’ which uses a Kalman Filter for track building or ‘Fit’ which uses a simple straight line fit for track building.
• full_track_info (bool) – If True predicted state vector of all DUTs is appended to track table in order to get full information on track. If False only state vector of DUT for which the track is fitted is appended to track table. This option is only possible if the Kalman Filter method is choosen.
• min_track_distance (iterable, boolean) – A minimum distance all track intersection at the DUT have to be apart, otherwise these tracks are deleted. This is needed to get a correct efficiency number, since assigning the same cluster to several tracks is error prone and will not be implemented. If it is true the std setting of 200 um is used. Otherwise a distance in um for each DUT has to be given. e.g.: For two devices: min_track_distance = (50, 250) If False, the minimum track distance is not considered.
• chunk_size (uint) – Chunk size of the data when reading from file.