<p>In <a href="https://github.com/mlpack/mlpack/pull/746#discussion_r76329137">src/mlpack/core/tree/cellbound_impl.hpp</a>:</p>
<pre style='color:#555'>&gt; +  // of hyperrectangles is restricted.
&gt; +  size_t numCorners = 0;
&gt; +  for (size_t pos = numEqualBits + 1; pos &lt; order * tmpHiAddress.n_elem; pos++)
&gt; +  {
&gt; +    size_t row = pos / order;
&gt; +    size_t bit = order - 1 - pos % order;
&gt; +
&gt; +    // This hyperrectangle is not contained entirely in the bound.
&gt; +    // So, the number of hyperrectangles should be increased.
&gt; +    if (tmpHiAddress[row] &amp; ((AddressElemType) 1 &lt;&lt; bit))
&gt; +      numCorners++;
&gt; +
&gt; +    // We ran out of the limit of hyperrectangles. In that case we enlare
&gt; +    // the last hyperrectangle.
&gt; +    if (numCorners &gt;= maxNumBounds / 2)
&gt; +      tmpHiAddress[row] |= ((AddressElemType) 1 &lt;&lt; bit);
</pre>
<p>I think the first strategy is best here.  I don't think it's a problem if the functions are not symmetric, and I think the performance impact will be negligible.  Definitely the first strategy is best, but if you want to make an extra pass that's fine too.  I just want to avoid a situation where a user says "but if I set maxNumBounds to 8 and 9 the performance is the same, and that doesn't make sense!"</p>

<p style="font-size:small;-webkit-text-size-adjust:none;color:#666;">&mdash;<br />You are receiving this because you are subscribed to this thread.<br />Reply to this email directly, <a href="https://github.com/mlpack/mlpack/pull/746/files/f17843fc8e8ef2b1c4b04d572c521575f20a1f3c#r76329137">view it on GitHub</a>, or <a href="https://github.com/notifications/unsubscribe-auth/AJ4bFDAX8_CzqxrZX9jsjfHAP14MwUgJks5qjgpcgaJpZM4JZrEi">mute the thread</a>.<img alt="" height="1" src="https://github.com/notifications/beacon/AJ4bFIDw4xvZuLlc2ZCQwuj8_b3jDf_Hks5qjgpcgaJpZM4JZrEi.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/746/files/f17843fc8e8ef2b1c4b04d572c521575f20a1f3c#r76329137"></link>
  <meta itemprop="name" content="View Pull Request"></meta>
</div>
<meta itemprop="description" content="View this Pull Request on GitHub"></meta>
</div>

<script type="application/json" data-scope="inboxmarkup">{"api_version":"1.0","publisher":{"api_key":"05dde50f1d1a384dd78767c55493e4bb","name":"GitHub"},"entity":{"external_key":"github/mlpack/mlpack","title":"mlpack/mlpack","subtitle":"GitHub repository","main_image_url":"https://cloud.githubusercontent.com/assets/143418/17495839/a5054eac-5d88-11e6-95fc-7290892c7bb5.png","avatar_image_url":"https://cloud.githubusercontent.com/assets/143418/15842166/7c72db34-2c0b-11e6-9aed-b52498112777.png","action":{"name":"Open in GitHub","url":"https://github.com/mlpack/mlpack"}},"updates":{"snippets":[{"icon":"PERSON","message":"@rcurtin in #746: I think the first strategy is best here.  I don't think it's a problem if the functions are not symmetric, and I think the performance impact will be negligible.  Definitely the first strategy is best, but if you want to make an extra pass that's fine too.  I just want to avoid a situation where a user says \"but if I set maxNumBounds to 8 and 9 the performance is the same, and that doesn't make sense!\""}],"action":{"name":"View Pull Request","url":"https://github.com/mlpack/mlpack/pull/746/files/f17843fc8e8ef2b1c4b04d572c521575f20a1f3c#r76329137"}}}</script>