Fix bug

yaeldMS · yaeldMS · commit c3df0e0a5808 · 2020-01-29T11:14:16.000+02:00
diff --git a/src/Microsoft.ML.Transforms/Dracula/Featurizer.cs b/src/Microsoft.ML.Transforms/Dracula/Featurizer.cs
@@ -3,7 +3,6 @@
 // See the LICENSE file in the project root for more information.
 
 using System;
-using System.Diagnostics;
 using System.Linq;
 using Microsoft.ML;
 using Microsoft.ML.Data;
@@ -157,12 +156,11 @@ public void GetFeatures(int iCol, int iSlot, Random rand, long key, Span<float>
                     countsBuffer[i++] = count;
             }
 
-            if (rand != null)
-                sum = AddLaplacianNoisePerLabel(iCol, rand, countsBuffer);
+            sum = AddLaplacianNoisePerLabel(iCol, rand, countsBuffer);
 
             // add log odds in the next _logOddsCount indices.
             GenerateLogOdds(iCol, countTable, countsBuffer, features.Slice(_labelBinCount, _logOddsCount), sum);
-            AssertValidOutput(features);
+            _host.Assert(FloatUtils.IsFinite(features));
 
             // Add the last feature: an indicator for isGarbage.
             features[NumFeatures - 1] = isGarbage ? 1 : 0;
@@ -173,12 +171,11 @@ public void GetFeatures(int iCol, int iSlot, Random rand, long key, Span<float>
         private float AddLaplacianNoisePerLabel(int iCol, Random rand, Span<float> counts)
         {
             _host.Assert(_labelBinCount == counts.Length);
-            _host.AssertValue(rand);
 
             float sum = 0;
             for (int ifeat = 0; ifeat < _labelBinCount; ifeat++)
             {
-                if (LaplaceScale[iCol] > 0)
+                if (rand != null && LaplaceScale[iCol] > 0)
                     counts[ifeat] += Stats.SampleFromLaplacian(rand, 0, LaplaceScale[iCol]);
 
                 // Clamp to zero when noise is too big and negative.
@@ -192,7 +189,7 @@ private float AddLaplacianNoisePerLabel(int iCol, Random rand, Span<float> count
         }
 
         // Fills _labelBinCount log odds features. One per class, or only one if 2 classes.
-        private void GenerateLogOdds(int iCol, ICountTable countTable, Span<float> counts, Span<float> logOdds, Single sum)
+        private void GenerateLogOdds(int iCol, ICountTable countTable, Span<float> counts, Span<float> logOdds, float sum)
         {
             _host.Assert(counts.Length == _labelBinCount);
             _host.Assert(logOdds.Length == _logOddsCount);
@@ -210,12 +207,5 @@ private void GenerateLogOdds(int iCol, ICountTable countTable, Span<float> count
                 }
             }
         }
-
-        [Conditional("DEBUG")]
-        private void AssertValidOutput(Span<float> features)
-        {
-            foreach (var feature in features)
-                _host.Assert(FloatUtils.IsFinite(feature));
-        }
     }
 }

Original file line number	Diff line number	Diff line change
`@@ -3,7 +3,6 @@`
`3`	`3`	`// See the LICENSE file in the project root for more information.`
`4`	`4`
`5`	`5`	`using System;`
`6`		`-using System.Diagnostics;`
`7`	`6`	`using System.Linq;`
`8`	`7`	`using Microsoft.ML;`
`9`	`8`	`using Microsoft.ML.Data;`
`@@ -157,12 +156,11 @@ public void GetFeatures(int iCol, int iSlot, Random rand, long key, Span<float>`
`157`	`156`	`countsBuffer[i++] = count;`
`158`	`157`	`}`
`159`	`158`
`160`		`- if (rand != null)`
`161`		`- sum = AddLaplacianNoisePerLabel(iCol, rand, countsBuffer);`
	`159`	`+ sum = AddLaplacianNoisePerLabel(iCol, rand, countsBuffer);`
`162`	`160`
`163`	`161`	`// add log odds in the next _logOddsCount indices.`
`164`	`162`	`GenerateLogOdds(iCol, countTable, countsBuffer, features.Slice(_labelBinCount, _logOddsCount), sum);`
`165`		`- AssertValidOutput(features);`
	`163`	`+ _host.Assert(FloatUtils.IsFinite(features));`
`166`	`164`
`167`	`165`	`// Add the last feature: an indicator for isGarbage.`
`168`	`166`	`features[NumFeatures - 1] = isGarbage ? 1 : 0;`
`@@ -173,12 +171,11 @@ public void GetFeatures(int iCol, int iSlot, Random rand, long key, Span<float>`
`173`	`171`	`private float AddLaplacianNoisePerLabel(int iCol, Random rand, Span<float> counts)`
`174`	`172`	`{`
`175`	`173`	`_host.Assert(_labelBinCount == counts.Length);`
`176`		`- _host.AssertValue(rand);`
`177`	`174`
`178`	`175`	`float sum = 0;`
`179`	`176`	`for (int ifeat = 0; ifeat < _labelBinCount; ifeat++)`
`180`	`177`	`{`
`181`		`- if (LaplaceScale[iCol] > 0)`
	`178`	`+ if (rand != null && LaplaceScale[iCol] > 0)`
`182`	`179`	`counts[ifeat] += Stats.SampleFromLaplacian(rand, 0, LaplaceScale[iCol]);`
`183`	`180`
`184`	`181`	`// Clamp to zero when noise is too big and negative.`
`@@ -192,7 +189,7 @@ private float AddLaplacianNoisePerLabel(int iCol, Random rand, Span<float> count`
`192`	`189`	`}`
`193`	`190`
`194`	`191`	`// Fills _labelBinCount log odds features. One per class, or only one if 2 classes.`
`195`		`- private void GenerateLogOdds(int iCol, ICountTable countTable, Span<float> counts, Span<float> logOdds, Single sum)`
	`192`	`+ private void GenerateLogOdds(int iCol, ICountTable countTable, Span<float> counts, Span<float> logOdds, float sum)`
`196`	`193`	`{`
`197`	`194`	`_host.Assert(counts.Length == _labelBinCount);`
`198`	`195`	`_host.Assert(logOdds.Length == _logOddsCount);`
`@@ -210,12 +207,5 @@ private void GenerateLogOdds(int iCol, ICountTable countTable, Span<float> count`
`210`	`207`	`}`
`211`	`208`	`}`
`212`	`209`	`}`
`213`		`-`
`214`		`- [Conditional("DEBUG")]`
`215`		`- private void AssertValidOutput(Span<float> features)`
`216`		`- {`
`217`		`- foreach (var feature in features)`
`218`		`- _host.Assert(FloatUtils.IsFinite(feature));`
`219`		`- }`
`220`	`210`	`}`
`221`	`211`	`}`