[MRG] MNT Initialize histograms in parallel and don't call np.zero in Hist-GBDT (#18341)

NicolasHug · web-flow · commit 4a2de5b86446 · 2020-09-04T15:48:39.000+02:00
diff --git a/doc/whats_new/v0.24.rst b/doc/whats_new/v0.24.rst
@@ -195,6 +195,13 @@ Changelog
   usage in `fit`. :pr:`18334` by `Olivier Grisel`_ `Nicolas Hug`_, `Thomas
   Fan`_ and `Andreas Müller`_.
 
+- |Efficiency| Histogram initialization is now done in parallel in
+  :class:`ensemble.HistGradientBoostingRegressor` and
+  :class:`ensemble.HistGradientBoostingClassifier` which results in speed
+  improvement for problems that build a lot of nodes on multicore machines.
+  :pr:`18341` by `Olivier Grisel`_, `Nicolas Hug`_, `Thomas Fan`_, and
+  :user:`Egor Smirnov <SmirnovEgorRu>`.
+
 - |API|: The parameter ``n_classes_`` is now deprecated in
   :class:`ensemble.GradientBoostingRegressor` and returns `1`.
   :pr:`17702` by :user:`Simona Maggio <simonamaggio>`.
diff --git a/sklearn/ensemble/_hist_gradient_boosting/histogram.pyx b/sklearn/ensemble/_hist_gradient_boosting/histogram.pyx
@@ -132,7 +132,8 @@ cdef class HistogramBuilder:
             G_H_DTYPE_C [::1] gradients = self.gradients
             G_H_DTYPE_C [::1] ordered_hessians = self.ordered_hessians
             G_H_DTYPE_C [::1] hessians = self.hessians
-            hist_struct [:, ::1] histograms = np.zeros(
+            # Histograms will be initialized to zero later within a prange
+            hist_struct [:, ::1] histograms = np.empty(
                 shape=(self.n_features, self.n_bins),
                 dtype=HISTOGRAM_DTYPE
             )
@@ -177,6 +178,12 @@ cdef class HistogramBuilder:
                 self.ordered_hessians[:n_samples]
             unsigned char hessians_are_constant = \
                 self.hessians_are_constant
+            unsigned int bin_idx = 0
+        
+        for bin_idx in range(self.n_bins):
+            histograms[feature_idx, bin_idx].sum_gradients = 0.
+            histograms[feature_idx, bin_idx].sum_hessians = 0.
+            histograms[feature_idx, bin_idx].count = 0
 
         if root_node:
             if hessians_are_constant:
@@ -227,7 +234,7 @@ cdef class HistogramBuilder:
         cdef:
             int feature_idx
             int n_features = self.n_features
-            hist_struct [:, ::1] histograms = np.zeros(
+            hist_struct [:, ::1] histograms = np.empty(
                 shape=(self.n_features, self.n_bins),
                 dtype=HISTOGRAM_DTYPE
             )
