joint_counts chi_2 failing

[sjsrey]

When testing esda against master in libpysal, one test is failing due to a change in https://github.com/pysal/libpysal/commit/9a303797dc68ad0db2c339ef54c2270a5950b5cc#diff-e8a3363e7d0110431da1549a5428007b having to do with what .w.to_adjlist(remove_symmetric=True) returns. Because the focal units that are now returned are different than before the change, what was a bw join becomes a wb and the contingency table changes from:

[[10, 4], [0, 10]] before the change, to:

[[10, 3], [1, 10]] after the change.

The fix will be to update the test in esda once a new version of libpysal gets released.