Decreasing is identical to Increasing in almost every way. It accepts the same parameters and produces the same outputs, with one obvious exception: it tries to isolate intervals of decrease rather than increase. This means that the gradient parameter is negative, not positive.