1. Test Modules
  2. Differential Validation
    1. Feedback Validation
    2. Learning Validation
    3. Total Accuracy
    4. Frozen and Alive Status
  3. Results

Target Description: The type Gaussian activation layer.

Report Description: The type Basic.

Subreport: Logs for com.simiacryptus.ref.lang.ReferenceCountingBase

Test Modules

Using Seed 3785945272054070272

Differential Validation

SingleDerivativeTester.java:153 executed in 0.00 seconds (0.000 gc):

        log.info(RefString.format("Inputs: %s", prettyPrint(inputPrototype)));
        log.info(RefString.format("Inputs Statistics: %s", printStats(inputPrototype)));
        log.info(RefString.format("Output: %s", outputPrototype.prettyPrint()));
        assert outputPrototype != null;
        log.info(RefString.format("Outputs Statistics: %s", outputPrototype.getScalarStatistics()));
      },
      outputPrototype.addRef(),
      RefUtil.addRef(inputPrototype)));
Logging
Inputs: [
[ [ 0.08 ], [ -0.128 ], [ -0.608 ] ],
[ [ 0.7 ], [ 0.496 ], [ 1.764 ] ]
]
Inputs Statistics: {meanExponent=-0.403119694464533, negative=2, min=-0.608, max=1.764, mean=0.38399999999999995, count=6, sum=2.304, positive=4, stdDev=0.747821725636086, zeros=0}
Output: [
[ [ 0.3976677055116089 ], [ 0.39568749504326983 ], [ 0.33161834545386687 ] ],
[ [ 0.31225393336676127 ], [ 0.35276733986812936 ], [ 0.08418095817883484 ] ]
]
Outputs Statistics: {meanExponent=-0.5525464417112985, negative=0, min=0.08418095817883484, max=0.3976677055116089, mean=0.3123626295704119, count=6, sum=1.8741757774224712, positive=6, stdDev=0.10668505000682163, zeros=0}

Feedback Validation

We validate the agreement between the implemented derivative of the inputs apply finite difference estimations:

SingleDerivativeTester.java:169 executed in 0.02 seconds (0.000 gc):

        return testFeedback(
            statistics,
            component.addRef(),
            RefUtil.addRef(inputPrototype),
            outputPrototype.addRef());
      },
      outputPrototype.addRef(),
      RefUtil.addRef(inputPrototype),
      component.addRef()));
Logging
Feedback for input 0
Inputs Values: [
[ [ 0.08 ], [ -0.128 ], [ -0.608 ] ],
[ [ 0.7 ], [ 0.496 ], [ 1.764 ] ]
]
Value Statistics: {meanExponent=-0.403119694464533, negative=2, min=-0.608, max=1.764, mean=0.38399999999999995, count=6, sum=2.304, positive=4, stdDev=0.747821725636086, zeros=0}
Implemented Feedback: [ [ -0.031813416440928714, 0.0, 0.0, 0.0, 0.0, 0.0 ], [ 0.0, -0.2185777533567329, 0.0, 0.0, 0.0, 0.0 ], [ 0.0, 0.0, 0.050647999365538536, 0.0, 0.0, 0.0 ], [ 0.0, 0.0, 0.0, -0.17497260057459216, 0.0, 0.0 ], [ 0.0, 0.0, 0.0, 0.0, 0.20162395403595104, 0.0 ], [ 0.0, 0.0, 0.0, 0.0, 0.0, -0.14849521022746467 ] ]
Implemented Statistics: {meanExponent=-0.9556661361758314, negative=4, min=-0.2185777533567329, max=0.20162395403595104, mean=-0.008932972977728581, count=36, sum=-0.3215870271982289, positive=2, stdDev=0.06276040785694621, zeros=30}
Measured Feedback: [ [ -0.03183317241362005, 0.0, 0.0, 0.0, 0.0, 0.0 ], [ 0.0, -0.2185857149172854, 0.0, 0.0, 0.0, 0.0 ], [ 0.0, 0.0, 0.05062853888626595, 0.0, 0.0, 0.0 ], [ 0.0, 0.0, 0.0, -0.1749858988181474, 0.0, 0.0 ], [ 0.0, 0.0, 0.0, 0.0, 0.20161350160252667, 0.0 ], [ 0.0, 0.0, 0.0, 0.0, 0.0, -0.1484863220253818 ] ]
Measured Statistics: {meanExponent=-0.9556489655000249, negative=4, min=-0.2185857149172854, max=0.20161350160252667, mean=-0.008934696324601168, count=36, sum=-0.32164906768564205, positive=2, stdDev=0.06276028788638831, zeros=30}
Feedback Error: [ [ -1.97559726913335E-5, 0.0, 0.0, 0.0, 0.0, 0.0 ], [ 0.0, -7.961560552494085E-6, 0.0, 0.0, 0.0, 0.0 ], [ 0.0, 0.0, -1.9460479272585818E-5, 0.0, 0.0, 0.0 ], [ 0.0, 0.0, 0.0, -1.3298243555254219E-5, 0.0, 0.0 ], [ 0.0, 0.0, 0.0, 0.0, -1.0452433424373853E-5, 0.0 ], [ 0.0, 0.0, 0.0, 0.0, 0.0, 8.888202082862895E-6 ] ]
Error Statistics: {meanExponent=-4.903720705913417, negative=5, min=-1.97559726913335E-5, max=8.888202082862895E-6, mean=-1.7233468725882939E-6, count=36, sum=-6.204048741317858E-5, positive=1, stdDev=5.503974976184558E-6, zeros=30}

Returns

    {
      "absoluteTol" : {
        "count" : 36,
        "sum" : 7.981689157890437E-5,
        "min" : 0.0,
        "max" : 1.97559726913335E-5,
        "sumOfSquare" : 1.1974919393421235E-9,
        "standardDeviation" : 5.324281499300601E-6,
        "average" : 2.217135877191788E-6
      },
      "relativeTol" : {
        "count" : 6,
        "sum" : 6.146141259686149E-4,
        "min" : 1.821186153529119E-5,
        "max" : 3.104011235618884E-4,
        "sumOfSquare" : 1.3661446942741475E-7,
        "standardDeviation" : 1.1079714856994842E-4,
        "average" : 1.0243568766143581E-4
      }
    }

Learning Validation

We validate the agreement between the implemented derivative of the internal weights apply finite difference estimations:

SingleDerivativeTester.java:185 executed in 0.00 seconds (0.000 gc):

        return testLearning(
            statistics,
            component.addRef(),
            RefUtil.addRef(inputPrototype),
            outputPrototype.addRef());
      },
      outputPrototype.addRef(),
      RefUtil.addRef(inputPrototype),
      component.addRef()));

Returns

    {
      "absoluteTol" : {
        "count" : 36,
        "sum" : 7.981689157890437E-5,
        "min" : 0.0,
        "max" : 1.97559726913335E-5,
        "sumOfSquare" : 1.1974919393421235E-9,
        "standardDeviation" : 5.324281499300601E-6,
        "average" : 2.217135877191788E-6
      },
      "relativeTol" : {
        "count" : 6,
        "sum" : 6.146141259686149E-4,
        "min" : 1.821186153529119E-5,
        "max" : 3.104011235618884E-4,
        "sumOfSquare" : 1.3661446942741475E-7,
        "standardDeviation" : 1.1079714856994842E-4,
        "average" : 1.0243568766143581E-4
      }
    }

Total Accuracy

The overall agreement accuracy between the implemented derivative and the finite difference estimations:

SingleDerivativeTester.java:200 executed in 0.00 seconds (0.000 gc):

    //log.info(String.format("Component: %s\nInputs: %s\noutput=%s", component, Arrays.toStream(inputPrototype), outputPrototype));
    log.info(RefString.format("Finite-Difference Derivative Accuracy:"));
    log.info(RefString.format("absoluteTol: %s", statistics.absoluteTol));
    log.info(RefString.format("relativeTol: %s", statistics.relativeTol));
Logging
Finite-Difference Derivative Accuracy:
absoluteTol: 2.2171e-06 +- 5.3243e-06 [0.0000e+00 - 1.9756e-05] (36#)
relativeTol: 1.0244e-04 +- 1.1080e-04 [1.8212e-05 - 3.1040e-04] (6#)

Frozen and Alive Status

SingleDerivativeTester.java:208 executed in 0.01 seconds (0.000 gc):

    testFrozen(component.addRef(), RefUtil.addRef(inputPrototype));
    testUnFrozen(component.addRef(), RefUtil.addRef(inputPrototype));

LayerTests.java:605 executed in 0.00 seconds (0.000 gc):

    throwException(exceptions.addRef());

Results

classdetailsresult
com.simiacryptus.mindseye.test.unit.SingleDerivativeTesterToleranceStatistics{absoluteTol=2.2171e-06 +- 5.3243e-06 [0.0000e+00 - 1.9756e-05] (36#), relativeTol=1.0244e-04 +- 1.1080e-04 [1.8212e-05 - 3.1040e-04] (6#)}OK
  {
    "result": "OK",
    "performance": {
      "execution_time": "0.167",
      "gc_time": "0.106"
    },
    "created_on": 1587005196683,
    "file_name": "derivativeTest",
    "report": {
      "simpleName": "Basic",
      "canonicalName": "com.simiacryptus.mindseye.layers.java.GaussianActivationLayerTest.Basic",
      "link": "https://github.com/SimiaCryptus/mindseye-java/tree/c9a1867488dc7e77a975f095285b5882c0486db6/src/test/java/com/simiacryptus/mindseye/layers/java/GaussianActivationLayerTest.java",
      "javaDoc": "The type Basic."
    },
    "archive": "s3://code.simiacrypt.us/tests/com/simiacryptus/mindseye/layers/java/GaussianActivationLayer/Basic/derivativeTest/202004164636",
    "id": "33646a1d-e1a7-4ca2-a522-4fa7d3864952",
    "report_type": "Components",
    "display_name": "Derivative Validation",
    "target": {
      "simpleName": "GaussianActivationLayer",
      "canonicalName": "com.simiacryptus.mindseye.layers.java.GaussianActivationLayer",
      "link": "https://github.com/SimiaCryptus/mindseye-java/tree/c9a1867488dc7e77a975f095285b5882c0486db6/src/main/java/com/simiacryptus/mindseye/layers/java/GaussianActivationLayer.java",
      "javaDoc": "The type Gaussian activation layer."
    }
  }