Filter

Athena bandpass filters data using two-way window-based FIR (Finite Impulse Response) filtering, which order is computed as 3 times the ratio between the sampling frequency and the minimum cut frequency of the frequency band in which the signal has to be filtered, rounded to the nearest integer towards zero.

This filter is zero-phase, so it has zero phase distortion.

Note that data is filtered in both the forward and reverse directions, so the resulting filter order is doubled.

This filter is based on a standard window-based MATLAB filter and a standard zero-phase digital MATLAB filtering process.

Athena allows also to utilize custom filters, whenever a filtering operation is needed (such as in extracting connectivity measures).

Every custom filter function's interface need to be in the following format:

  smoothdata = custom_filter(data, fs, fmin, fmax)

Where:

• data is the (channels x time samples) data matrix
• fs is the sample frequency
• fmin is the lower cut frequency of the band to extract
• fmax is the higher cut frequency
• smoothdata is the filtered data matrix (channels x time samples)

The file which contains the filtering function needs to have the same name of the function itself.