HRFBold

observations.tvb_monitors.HRFBold(
    k_1=5.6,
    V_0=0.02,
    period=1000.0,
    downsample_period=4.0,
    kernel=None,
    downsample=None,
    voi=None,
    convolution_mode='valid',
    history=None,
)

BOLD signal monitor using hemodynamic response function convolution.

This monitor simulates the Blood Oxygen Level Dependent (BOLD) signal by: 1. Downsampling the neural activity 2. Convolving with a hemodynamic response function kernel 3. Downsampling to the final BOLD sampling period

Calling the monitor applies it to a stored solution. Passing it through a native solver’s observe= keyword computes the same valid causal convolution block-wise; online execution requires convolution_mode="valid" and a downsampler implementing the streaming preparation contract. Any callable downsampler is supported post-hoc. Downsampler state is carried between blocks and its live parameters are available under config.monitor.downsample. By default the convolution input is a TemporalAverage at downsample_period. Pass downsample=Identity() to use the actual incoming grid without another averaging stage.

Warm history may be an already-downsampled array or a solution on the monitor’s pre-downsampling input grid. In a joint branch, that input is the shared stage’s output; raw history is not replayed through the shared stage automatically.

Methods

Name	Description
prepare	Prepare causal valid-mode HRF convolution for block-wise execution.

prepare

observations.tvb_monitors.HRFBold.prepare(grid, sample, variable_names)

Prepare causal valid-mode HRF convolution for block-wise execution.