Signal processing

These functions work on plain numpy arrays and do not keep any state. They are in the submodule pp.signal (e.g., pp.signal.smooth_window). The methods of the data classes call them with the right arguments, so you usually do not need them directly, but they are useful for your own processing steps or for data that are not in a data object.

Filtering and baseline

butter_lowpass	Get lowpass-filter coefficients for Butterworth-filter.
butter_lowpass_filter	Lowpass-filter signal using a Butterworth-filter.
lowpass_filter_iterative	Lowpass filter that handles NaN values via iterative interpolation.
smooth_window	Smooth the data using a window with requested size.
downsample	Simple downsampling scheme using mean within the downsampling window.
zscale	Z-transform an array, ignoring NaNs.
bspline	Re-implementation from https://mc-stan.org/users/documentation/case-studies/splines_in_stan.html.
baseline_envelope	Extract baseline based on the lower envelope of the (filtered) signal.
baseline_envelope_iter_bspline	Extract baseline based on (re-)estimating the lower envelope using B-splines.
baseline_pupil_model	Extract baseline based on filtering after removing stim-locked activity.

Blinks

velocity_savgol	Compute velocity (first derivative) using Savitzky-Golay filter.
detect_blinks_velocity	Detect blinks as everything between a fast downward and a fast upward-trending PD-changes.
detect_blinks_zero	Detect blinks as consecutive sequence of blink_val (e.g., 0 or NaN) of at least min_duration successive values in the signal sy.
blink_onsets_mahot	Method for finding the on- and offset for each blink (excluding transient).

Pupil response

pupil_kernel	According to Hoeks and Levelt (1993, https://link.springer.com/content/pdf/10.3758%2FBF03204445.pdf), the PRF can be described by the so-called Erlang gamma function $$h_{HL}(t) = t^n e^{-nt/t_{max}}$$ which we normalize to $$h(t)=frac{1}{h_{max}} h_{HL}(t)$$ where $$h_{max} = max_t{left(h_{HL}(t)right)} = e^{-n}t_{max}^{n}$$ which yields a maximum value of 1 for this function.
pupil_kernel_t	According to Hoeks and Levelt (1993, https://link.springer.com/content/pdf/10.3758%2FBF03204445.pdf), the PRF can be described by the so-called Erlang gamma function $$h_{HL}(t) = t^n e^{-nt/t_{max}}$$ which we normalize to $$h(t)=frac{1}{h_{max}} h_{HL}(t)$$ where $$h_{max} = max_t{left(h_{HL}(t)right)} = e^{-n}t_{max}^{n}$$ which yields a maximum value of 1 for this function.
pupil_get_max_duration	Get the time when the PRF with parameters $n$ and $t_{max}$ is decayed to thr.
plot_prf	Plot profile of the pupil-response function (PRF) with parameters npar and tmax.
pupil_build_design_matrix	Construct design matrix (nevents x n).
pupil_response	Estimate pupil-response based on event-onsets.
pupil_response_coverage	Share of each event's response that falls on valid samples.
pupilresponse_nnls	Estimate single-event pupil responses based on canonical PRF (pupil_kernel()) using non-negative least-squares (NNLS).
pupil_signal_quality	Compute local signal quality metric for pupil signal.
pupil_eye_offset	Compute a slowly-varying offset between left and right pupil signals.

Simulated data

fake_pupil_baseline	Generate smooth baseline pupil size via filtered noise.
fake_gaze_fixations	Generate step-function gaze positions (fixations with random locations and durations).
add_measurement_noise	Add Gaussian measurement noise to a signal.
generate_foreshortening_data	Generate complete synthetic dataset for testing foreshortening correction algorithm.