# dbcsp 0.0.2.1

## Bug fixes

• Warning message changed. If the minimum eigenvalue is below the tolerance indicated by eig.tol when creating the dbcsp object, average covariance matrices are replaced by the most similar matrix that is positive definite and a warning message is printed to make the user aware of it.

# dbcsp 0.0.2.0

## New features

• Missing values handling is available. The missing values along a signal are imputed by a linear interpolation.

• User-defined custom distances are allowed. A function $$d(x_1,x_2)$$ which returns a scalar providing the distance value between $$x_1, x_2$$ can be used as distance function when creating the dbcsp object. The name of the function has to be passed to type parameter.

• New parameter getsignals in plot(). If getsignals=TRUE the projected signals for a given instance within a given class are returned, that is, the plotted projected signals are returned.

## Bug fixes

• Optional arguments of matplotlib function can be passed to plot() function.

• In plot() the original signals are shown before standardizing.

• The dbcsp::train function was being masked by caret::train function. This has been fixed.

# dbcsp 0.0.1.0

First public release.