tlby commented on issue #14461: [MXNET-1359] Adds a multiclass-MCC metric
derived from Pearson
URL: https://github.com/apache/incubator-mxnet/pull/14461#issuecomment-477411935
It turned out building the confusion matrix wasn't the hotspot in
`.update()` anyway. Once I cleaned up the actua
tlby commented on issue #14461: [MXNET-1359] Adds a multiclass-MCC metric
derived from Pearson
URL: https://github.com/apache/incubator-mxnet/pull/14461#issuecomment-475484260
```diff
diff --git a/python/mxnet/metric.py b/python/mxnet/metric.py
index 2a33cf4d9..7bc090a0a 100644
---
tlby commented on issue #14461: [MXNET-1359] Adds a multiclass-MCC metric
derived from Pearson
URL: https://github.com/apache/incubator-mxnet/pull/14461#issuecomment-475331531
I think this PR is ready now, sorry for posting the PR prematurely. I am
happy with test coverage at this point, a
tlby commented on issue #14461: [MXNET-1359] Adds a multiclass-MCC metric
derived from Pearson
URL: https://github.com/apache/incubator-mxnet/pull/14461#issuecomment-475067430
```diff
diff --git a/python/mxnet/metric.py b/python/mxnet/metric.py
index 2a33cf4d9..6de76cc64 100644
---
tlby commented on issue #14461: [MXNET-1359] Adds a multiclass-MCC metric
derived from Pearson
URL: https://github.com/apache/incubator-mxnet/pull/14461#issuecomment-474147085
hmm, actually `numpy.corrcoef()` may not be working how I need for the
multiclass class. This work is based on th