Source code for allensdk.brain_observatory.behavior.trial_masks

import numpy as np
import pandas as pd


[docs]def trial_types(trials, trial_types):
    """ only include trials of certain trial types

    Parameters
    ----------
    trials : pandas DataFrame
        dataframe of trials
    trial_types : list or other iterator


    Returns
    -------
    mask : pandas Series of booleans, indexed to trials DataFrame

    """

    if trial_types is not None and len(trial_types) > 0:
        return trials['trial_type'].isin(trial_types)
    else:
        return pd.Series(np.ones((len(trials), ), dtype=bool), 
                         name="trial_type", index=trials.index)




[docs]def contingent_trials(trials):
    """ GO & CATCH trials only

    Parameters
    ----------
    trials : pandas DataFrame
        dataframe of trials

    Returns
    -------
    mask : pandas Series of booleans, indexed to trials DataFrame

    """
    return trial_types(trials, ('go', 'catch'))




[docs]def reward_rate(trials, thresh=2.0):
    """ masks trials where the reward rate (per minute) is below some threshold.

    This de facto omits trials in which the animal was not licking for extended periods
    or periods when they were licking indiscriminantly.

    Parameters
    ----------
    trials : pandas DataFrame
        dataframe of trials
    thresh : float, optional
        threshold under which trials will not be included, default: 2.0

    Returns
    -------
    mask : pandas Series of booleans, indexed to trials DataFrame

    """

    mask = trials['reward_rate'] > thresh
    return mask