utils module

Collection of different functions and tools for reading data from different tasks and tools for data analysis @Author: Mehrdad Kashefi

class utils.Analysis_tools(fs)

Class for tools often required for data analysis

align_continuous(trial_info, D, condition_columns, align_column, channel_column, t_range, return_mean=True)

Aligns continuous data on a specific event

Parameters:

  • trial_info (df) – Data frame of the trial information (Start time, end time, etc.)

  • D (df) – Data frame of the continuous data (Hand pos, Hand vel, etc.)

  • condition_columns (list) – Columns to group the data

  • align_column (str) – Column to align the data on

  • channel_column (str) – Which data column is selected

  • t_range (list) – Time range to align the data

  • return_mean (bool) – Whether to return the mean of trials of a condition, default is True

Returns:

  • data_aligned (np.array) – Aligned data (condition x time x channel)

  • conds (np.array) – Unique conditions

  • aligned_trials_type (np.array) – Type of the trials

align_fr(trial_info, units, condition_columns, align_column, unit, t_range, return_mean=True)

Aligns units firing rate on a specific event

Parameters:

  • trial_info (df) – Data frame of the trial information (Start time, end time, etc.)

  • units (df) – Data frame of the units (Spike times, KS label, etc.)

  • condition_columns (list) – Columns to group the data

  • align_column (str) – Column to align the data on

  • unit (list) – Which unit to get the spike times

  • t_range (list) – Time range to align the data

  • return_mean (bool) – Whether to return the mean of trials of a condition, default is True

Returns:

data_aligned (np.array) – Aligned data (condition x time x unit)

align_stem(trial_info, units, condition_columns, align_column, unit, t_range)

Aligns units firing rate on a specific event

Parameters:

  • trial_info (df) – Data frame of the trial information (Start time, end time, etc.)

  • units (df) – Data frame of the units (Spike times, KS label, etc.)

  • condition_columns (list) – Columns to group the data

  • align_column (str) – Column to align the data on

  • unit (list) – Which unit to get the spike times

  • t_range (list) – Time range to align the data

Returns:

data_aligned (np.array) – Aligned data (condition x time x unit)

get_fr(which_unit, units, t, **kwargs)

Get firing rate for a specific unit, during a specific time

Parameters:

  • which_unit (int) – Which unit to get the spike times

  • units (df) – Data frame of the units (Spike times, KS label, etc.)

  • t (list) – Time range to get the spike times

  • gauss_width (int) – Window size for the Gaussian window, default is 50

  • dt (float) – Step size, default is 1/1000

  • t_pad (float) – How many seconds to pad the data from before and after the trial, default is 0.5

  • plot (bool) – Plot the single trial, default is False

  • return_padded (bool) – Whether to include padded data in return, default is False

  • causal (bool) – Whether to use causal convolution, default is True

Returns:

FR (np.array) – Continuous firing rate for the specific unit, in the specific time range

get_stem(units, which_unit, t, **kwargs)

Get spike times for a specific unit, in a specific time range

Parameters:

  • units (df) – Data frame of the units (Spike times, KS label, etc.)

  • which_unit (int) – Which unit to get the spike times

  • t (list) – Time range to get the spike times

  • df (float) – Step size, default is 1/1000

  • t_pad (float) – How many seconds to pad the data from before and after the trial, default is 0.5

  • plot (bool) – Plot the single trial, default is False

  • return_padded (bool) – Whether to include padded data in return, default is False

Returns:

stem (np.array) – Spike times for the specific unit, in the specific time range

utils.collapse_cond_time(data)

Collapse the condition and time dimensions of the data. :param data: numpy array of shape (n_cond, n_time, n_chan)

Returns:

numpy array of shape (n_time * n_cond, n_chan)

utils.cosine_similarity(D1, D2)

Calculate the cosine similarity between two matrices. :param D1: First matrix :type D1: np.array :param D2: Second matrix :type D2: np.array

Returns:

float – Cosine similarity between the two matrices

utils.dist_to_G(dist)

Transforms a squared Euclidean matrix into a second moment matrix :param dist: 2d array :type dist: ndarray

utils.get_G(data, normalize=False)

Get the second moment matrix from the data :param data: 2D array of shape (K, N) where K is the number of conditions and N is the number of samples :type data: np.array :param normalize: If True, normalize the second moment matrix by its trace :type normalize: bool

Returns:

np.array – Second moment matrix of shape (K, K)

utils.get_condition_mean(D, indicator)

Get the mean of the conditions

Parameters:

  • D (np.array) – Data matrix

  • indicator (np.array) – Indicator for the conditions,

utils.get_indicator_matrix(data)

Get the indicator matrix for the data

Parameters:

data (np.array) – One dimensional data contianing the conditions

Returns:

indicator_matrix (np.array) – Indicator matrix

utils.read_mc_maze(data_path)

Read the data from the NHP maze task

Parameters:

data_path (str) – Path to the data set

Returns:

  • D (df) – Data frame of the continuous data (Hand pos, Hand vel, etc.)

  • trial_info (df) – Data frame of the trial information (Start time, end time, etc.)

  • units (df) – Data frame of the units (Spike times, KS label, etc.)

utils.read_mc_rtt(data_path)

Read the data from the NHP random target reaching task

Parameters:

data_path (str) – Path to the data set

Returns:

  • D (df) – Data frame of the continuous data (Hand pos, Hand vel, etc.)

  • trial_info (df) – Data frame of the trial information (Start time, end time, etc.)

  • units (df) – Data frame of the units (Spike times, KS label, etc.)

utils.read_nhp_sequence(data_path, **kwargs)

Read the data from the NHP sequence task

Parameters:

data_path (str) – Path to the data set

Returns:

  • D (df) – Data frame of the continuous data (Hand pos, Hand vel, etc.)

  • trial_info (df) – Data frame of the trial information (Start time, end time, etc.)

  • units (df) – Data frame of the units (Spike times, KS label, etc.)

utils.read_point2point(data_path, **kwargs)

Read the data from point2point task

Parameters:

data_path (str) – Path to the data set

Returns:

  • D (df) – Data frame of the continuous data (Hand pos, Hand vel, etc.)

  • trial_info (df) – Data frame of the trial information (Start time, end time, etc.)

  • units (df) – Data frame of the units (Spike times, KS label, etc.)

utils.resample(Data, timestamps)

Resamples using a linear interpolation

Parameters:

  • Data (np.array) – Data to be resampled

  • timestamps (np.array) – Timestamps for the data

Returns:

resampled_data (np.array)