New issue 1215: Constructing AMRKDTree should fail early for oct and particle codes https://bitbucket.org/yt_analysis/yt/issues/1215/constructing-amrkdtree-shou... Nathan Goldbaum: Currently the AMRKDTree will seg on construction for oct and particle codes (see #1212). We should fail early instead of seg faulting. We should also mention in the documentation that this is an issue. In the long term we will avoid constructing a KDTree at all for these codes.