<p>I prefer the first version. In fact, I've implemented the negative log likelihood for another model, which also requires the error value, so now there isn't just one function that requires an interface modification.</p>

<p style="font-size:small;-webkit-text-size-adjust:none;color:#666;">&mdash;<br>Reply to this email directly or <a href="https://github.com/mlpack/mlpack/pull/451#issuecomment-163740664">view it on GitHub</a>.<img alt="" height="1" src="https://github.com/notifications/beacon/AJ4bFHoR-jQGX8bvl_5GAOyJAE9Rmb0cks5pOdgRgaJpZM4GAqt4.gif" width="1" /></p>
<div itemscope itemtype="http://schema.org/EmailMessage">
<div itemprop="action" itemscope itemtype="http://schema.org/ViewAction">
  <link itemprop="url" href="https://github.com/mlpack/mlpack/pull/451#issuecomment-163740664"></link>
  <meta itemprop="name" content="View Pull Request"></meta>
</div>
<meta itemprop="description" content="View this Pull Request on GitHub"></meta>
</div>