eden-kramer-lab/ripple_detection

Tools for identifying sharp wave ripple events using LFPs

Eden-Kramer-Lab

Website

ripple_detection's Issues

Add ripple size metric

maximum threshold ( in units of sd) at which the event would still be detected (ie be above for >15ms or whatever your time threshold was)

Issue with filter_ripple_band padlen

Hello,

I wanted to try playing around with my own LFP ephys recordings, but cannot seem to run my lfps array (shape=(7350,)) with the filter_ripple_band(lfps) function to get the filtered_lfps.

The ValueError I get is "The length of the input vector x must be greater than padlen, which is 954." which is traced back to the "filtered_data[~is_nan] = filtfilt(filter_numerator, filter_denominator, data[~is_nan], axis=0)" section of the code.

Let me know if you have any ideas or need any more information!

80-125Hz ripple detection

Dear edeno,
I'm sorry to bother you.
Does this code apply to ripple detection at 80-125Hz?
Looking forward to your reply

Yours,
Xiaoke

filter_ripple_band assumes there are nans in lfp array and throws an error if there aren't any.

Line 89 in ripple_detection/blob/master/ripple_detection/core.py is throwing an error since there are no nans in my lfp data.

def filter_ripple_band(data):
    '''Returns a bandpass filtered signal between 150-250 Hz
    Parameters
    ----------
    data : array_like, shape (n_time,)
    Returns
    -------
    filtered_data : array_like, shape (n_time,)
    '''
    filter_numerator, filter_denominator = _get_ripplefilter_kernel()
    is_nan = np.any(np.isnan(data), axis=-1)
    filtered_data = np.full_like(data, np.nan)
    filtered_data[~is_nan] = filtfilt(
        filter_numerator, filter_denominator, data[~is_nan], axis=0)
    return filtered_data

I suggest the following change:

def filter_ripple_band(data):
    '''Returns a bandpass filtered signal between 150-250 Hz
    Parameters
    ----------
    data : array_like, shape (n_time,)
    Returns
    -------
    filtered_data : array_like, shape (n_time,)
    '''
    filter_numerator, filter_denominator = _get_ripplefilter_kernel()
    is_nan = np.any(np.isnan(data), axis=-1)
    if is_nan==False:
        filtered_data = filtfilt(filter_numerator, filter_denominator, data, axis=0)
    else:
        filtered_data = np.full_like(data, np.nan)
        filtered_data[~is_nan] = filtfilt(
        filter_numerator, filter_denominator, data[~is_nan], axis=0)

    return filtered_data

All detectors are giving axis out of bounds errors

Where I supply LFP data as a nd array or as a 1d array all the dectors produce the following error: AxisError: axis 1 is out of bounds for array of dimension 1

add ripple amplitude to output table

in line 117 of ripple_detection/ripple_detection/detectors.py, if you can add ripple_amplitude to the ripple_combined variable and then add all three columns in the return function on line 122, that would be great. thanks.

Reading raw neuropixels data into python

I am working with some raw neuropixels LFP data, its compressed into a .tar but I am not sure how to open it in python.