janelia_core.dataprocessing.ahrens_lab
Data tools used for working with results from the Ahrens lab.
While the tools here were originally developed for working with Ahrens lab data, they are written to be generally useful for many datasets.
Module Contents
Functions
|
A function to take raw ephys data and down-sample it to the imaging rate of an experiment. |
|
Finds indices where the camera was triggered in the raw ephys data. |
- janelia_core.dataprocessing.ahrens_lab.down_sample_ephys_vls(ts: numpy.ndarray, image_acq_starts: numpy.ndarray, n_images: int, vls: numpy.ndarray, ds_type: str = 'mean', acq_start_type: str = 'boolean') Tuple[numpy.ndarray, numpy.ndarray]
A function to take raw ephys data and down-sample it to the imaging rate of an experiment.
Values are down-sampled between image acquisitions, so that the down-sampled value for image i will be a function of the data that falls between the start of acquisition of image i and the start of acquisition for image i+1.
- Args:
ts: Time stamps for the raw ephys data.
image_acq_starts: An array indicating the time image acquisition starts for each volume. This can be either a binary array of an array of time stamps (see acq_start_type below for more details).
n_images: The number of images that were acquired.
vls: The raw ephys values to down sample
ds_type: The type of down-sampling to perform. Options are:
‘mean’ The down-sampled value is the mean value between the start of image acquisitions.
‘mode’ The down-sampled value is the mode between the start of image acquisitions.
‘max’ The down-sampled value is the max between the start of image acquisitions.
‘consant’ The down-sampled value is the constant value between the start of image acquisitions. If selected, a check will be performed to ensure values are actually constant, and if this check is failed, an error will be raised.
acq_start_type: Indicates the way of specifying image acquisition starts. If ‘boolean’, then image_acq_starts should be a binary array with a value of 1 indicating each point in time image acquisition starts. If ‘time’ then this should be an array with times that image acqusition starts.
- Returns:
ds_ts: The time stamps of the down sampled values. Time stamps correspond to the start of image acquisitions.
ds_vls: The down sampled values.
- Raises:
ValueError: If n_images is greater than the number of image starts in image_acq_starts. We require n_images to be strictly less than the number of image acquisition starts so a period of image acquisition is well defined for the last requested image.
ValueError: If the first or last image acquisition times fall outside of the range of time stamps for the values.
ValueError: If ‘constant’ down-sampling is selected, but non-constant values are detected in a sampling period.
- janelia_core.dataprocessing.ahrens_lab.find_camera_triggers_in_ephys(sig: numpy.ndarray, th: float = 3.8, smp_tol: int = 2) numpy.ndarray
Finds indices where the camera was triggered in the raw ephys data.
- Args:
sig: 1-d array of camera trigger data
th: the threshold that indicates a camera threshold
smp_tol: After extracting camera triggers, we do a check and make sure all camera triggers are the same number of samples apart +/- a tolerance; this is the tolerance
- Returns:
inds: The indices of sig where values first pass th
- Raises:
ValueError: If the first value of sig is greather than th (since then timing of the onset cannot be determined)
RuntimeError: If the tolerance check is not passed