stytra package

Subpackages

Submodules

stytra.utilities module

class stytra.utilities.Database[source]

Bases: object

inset_experiment_data(exp_data)[source]
Parameters:exp_data (the data collector dictionary) –
Returns:
Return type:index of database entry
class stytra.utilities.FramerateRecorder(n_fps_frames=5)[source]

Bases: object

update_framerate()[source]

Calculate the framerate every n_fps_frames frames.

class stytra.utilities.FrameProcess(name='', n_fps_frames=10)[source]

Bases: multiprocessing.context.Process

A basic class for a process that deals with frames. It provides framerate calculation.

Parameters:n_fps_frames – the maximal number of frames to use to calculate framerate
update_framerate()[source]
stytra.utilities.prepare_json(it, **kwargs)[source]

Used to create a dictionary which will be safe to put in MongoDB

Parameters:
  • it – the item which will be recursively sanitized
  • **kwargs –
    convert_datetime: bool
    if datetiems are to be converted to strings for JSON serialization
    eliminate_df: bool
    remove dataframes from the dictionary
stytra.utilities.get_default_args(func)[source]

Find default arguments of functions

Parameters:func –
stytra.utilities.strip_values(it)[source]
Parameters:it –
stytra.utilities.interpolate_nan(a)[source]
stytra.utilities.get_classes_from_module(input_module, parent_class)[source]

Find all the classes in a module that are children of a parent one.

Parameters:
  • input_module – module object
  • parent_class – parent class object
Returns:

OrderedDict of subclasses found
Return type:

type
stytra.utilities.recursive_update(d, u)[source]

Simple recursive update of dictionaries, from StackOverflow

Parameters:
  • d – dict to update
  • u – new values
Returns:
stytra.utilities.reduce_to_pi[source]

Puts an angle or array of angles inside the (-pi, pi) range

Module contents

class stytra.Stytra(recording=None, exec=True, **kwargs)[source]

Bases: object

Stytra application instance. Contains the QApplication and constructs the appropriate experiment object for the specified parameters

Parameters:
  • protocol (Protocol) – the protocols to be made available from the dropdown
  • display (dict) –
    full_screen : bool
    displays the stimulus full screen on the secondary monitor, otherwise it is in a window
    window_size : tuple(int, int)
    optional specification of the size of the stimulus display area
  • camera (dict) –
    video_file : str
    or
    type: str
    supported cameras are “ximea” (with the official API) “avt” (With the Pymba API) “spinnaker” (PointGray/FLIR) “mikrotron” (via NI Vision C API)
    rotation: int
    how many times to rotate the camera image by 90 degrees to get the right orientation, matching the projector
    downsampling: int
    how many times to downsample the image (for some ximea cameras)
  • tracking (dict) –
    preprocessing_method: str, optional
    ”prefilter” or “bgsub”
    method: str
    one of “tail”, “eyes” or “fish”
    estimator: str or class
    for closed-loop experiments: either “vigor” for embedded experiments
    or “position” for freely-swimming ones. A custom estimator can be supplied.
  • recording (bool) – for video-recording experiments
  • embedded (bool) – if not embedded, use circle calibrator to match the camera and projector
  • dir_assets (str) – the location of assets used for stimulation (pictures, videos, models for closed loop etc.)
  • dir_save (str) – directory where the experiment data will be saved
  • metadata_animal (class) – subclass of AnimalMetadata adding information from a specific lab (species, genetic lines, pharmacological treatments etc.)
  • metadata_general (class) – subclass of GeneralMetadata, containing lab-specific information (setup names, experimenter names…)
  • record_stim_framerate (int) – if non-0 recodrds the displayed stimuli into an array which is saved alongside the other data.
  • trigger (object) – a trigger object, synchronising stimulus presentation to imaging acquisition
  • n_tracking_processes (int) – number of tracking processes to be used. Using more than 1 can improve performance but also cause issues in state-dependent tracking functions.