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

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

Test Modules

Using Seed 9165671521084806144

Differential Validation

SingleDerivativeTester.java:101 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: [
[ [ -0.852, 1.108, -1.616 ], [ -1.688, -0.712, 1.512 ] ],
[ [ 1.912, 0.028, 1.356 ], [ -0.804, 1.048, 1.556 ] ]
]
Inputs Statistics: {meanExponent=-0.04822021070395558, negative=5, min=-1.688, max=1.912, mean=0.23733333333333337, count=12, sum=2.8480000000000003, positive=7, stdDev=1.2646689509731612, zeros=0}
Output: [
[ [ -2.572, 0.08000000000000007, -2.0 ], [ -3.408, -1.74, 1.1280000000000001 ] ],
[ [ 0.19199999999999995, -1.0, 0.9720000000000001 ], [ -2.524, 0.020000000000000018, 1.1720000000000002 ] ]
]
Outputs Statistics: {meanExponent=-0.12643638895994422, negative=6, min=-3.408, max=1.1720000000000002, mean=-0.8066666666666668, count=12, sum=-9.680000000000001, positive=6, stdDev=1.539851363678398, zeros=0}

Feedback Validation

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

SingleDerivativeTester.java:117 executed in 0.04 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: [
[ [ -0.852, 1.108, -1.616 ], [ -1.688, -0.712, 1.512 ] ],
[ [ 1.912, 0.028, 1.356 ], [ -0.804, 1.048, 1.556 ] ]
]
Value Statistics: {meanExponent=-0.04822021070395558, negative=5, min=-1.688, max=1.912, mean=0.23733333333333337, count=12, sum=2.8480000000000003, positive=7, stdDev=1.2646689509731612, zeros=0}
Implemented Feedback: [ [ 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, ... ], ... ]
Implemented Statistics: {meanExponent=0.0, negative=0, min=0.0, max=1.0, mean=0.08333333333333333, count=144, sum=12.0, positive=12, stdDev=0.2763853991962833, zeros=132}
Measured Feedback: [ [ 1.0000000000021103, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.9999999999998899, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 1.0000000000021103, 0.0, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.9999999999976694, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.9999999999998899, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.0, 0.9999999999998899, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.9999999999998899, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.9999999999998899, ... ], ... ]
Measured Statistics: {meanExponent=-2.0855188676714649E-13, negative=0, min=0.0, max=1.0000000000021103, mean=0.08333333333329332, count=144, sum=11.999999999994237, positive=12, stdDev=0.27638539919615057, zeros=132}
Feedback Error: [ [ 2.1103119252074976E-12, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, -1.1013412404281553E-13, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 2.1103119252074976E-12, 0.0, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, -2.3305801732931286E-12, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, -1.1013412404281553E-13, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.0, -1.1013412404281553E-13, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, -1.1013412404281553E-13, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, -1.1013412404281553E-13, ... ], ... ]
Error Statistics: {meanExponent=-12.302493257665207, negative=10, min=-2.3305801732931286E-12, max=2.1103119252074976E-12, mean=-4.001737213204453E-14, count=144, sum=-5.7625015870144125E-12, positive=2, stdDev=4.60037933227515E-13, zeros=132}

Returns

    {
      "absoluteTol" : {
        "count" : 144,
        "sum" : 1.4203749287844403E-11,
        "min" : 0.0,
        "max" : 2.3305801732931286E-12,
        "sumOfSquare" : 3.0706025771606137E-23,
        "standardDeviation" : 4.5111750481234543E-13,
        "average" : 9.86371478322528E-14
      },
      "relativeTol" : {
        "count" : 12,
        "sum" : 7.1018746439254245E-12,
        "min" : 5.50670620214108E-14,
        "max" : 1.1652900866479222E-12,
        "sumOfSquare" : 7.676506442909495E-24,
        "standardDeviation" : 5.380097960751264E-13,
        "average" : 5.918228869937854E-13
      }
    }

Learning Validation

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

SingleDerivativeTester.java:133 executed in 0.01 seconds (0.000 gc):

        return testLearning(
            statistics,
            component.addRef(),
            RefUtil.addRef(inputPrototype),
            outputPrototype.addRef());
      },
      outputPrototype.addRef(),
      RefUtil.addRef(inputPrototype),
      component.addRef()));
Logging 
Learning Gradient for weight setByCoord 0
Weights: [ -1.72, -1.028, -0.384 ]
Implemented Gradient: [ [ 1.0, 1.0, 1.0, 1.0, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, 1.0, 1.0, 1.0, 1.0, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ... ] ]
Implemented Statistics: {meanExponent=0.0, negative=0, min=0.0, max=1.0, mean=0.3333333333333333, count=36, sum=12.0, positive=12, stdDev=0.4714045207910317, zeros=24}
Measured Gradient: [ [ 1.0000000000021103, 0.9999999999998899, 1.0000000000021103, 0.9999999999976694, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.9999999999998899, 0.9999999999998899, 0.9999999999998899, 0.9999999999998899, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ... ] ]
Measured Statistics: {meanExponent=-2.0855188676714649E-13, negative=0, min=0.0, max=1.0000000000021103, mean=0.3333333333331733, count=36, sum=11.999999999994237, positive=12, stdDev=0.4714045207908053, zeros=24}
Gradient Error: [ [ 2.1103119252074976E-12, -1.1013412404281553E-13, 2.1103119252074976E-12, -2.3305801732931286E-12, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, -1.1013412404281553E-13, -1.1013412404281553E-13, -1.1013412404281553E-13, -1.1013412404281553E-13, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ... ] ]
Error Statistics: {meanExponent=-12.302493257665207, negative=10, min=-2.3305801732931286E-12, max=2.1103119252074976E-12, mean=-1.6006948852817813E-13, count=36, sum=-5.7625015870144125E-12, positive=2, stdDev=9.095729322955472E-13, zeros=24}

Returns

    {
      "absoluteTol" : {
        "count" : 180,
        "sum" : 2.8407498575688805E-11,
        "min" : 0.0,
        "max" : 2.3305801732931286E-12,
        "sumOfSquare" : 6.141205154321227E-23,
        "standardDeviation" : 5.623798445728687E-13,
        "average" : 1.5781943653160448E-13
      },
      "relativeTol" : {
        "count" : 24,
        "sum" : 1.4203749287850849E-11,
        "min" : 5.50670620214108E-14,
        "max" : 1.1652900866479222E-12,
        "sumOfSquare" : 1.535301288581899E-23,
        "standardDeviation" : 5.380097960751264E-13,
        "average" : 5.918228869937854E-13
      }
    }

Total Accuracy

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

SingleDerivativeTester.java:148 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: 1.5782e-13 +- 5.6238e-13 [0.0000e+00 - 2.3306e-12] (180#)
relativeTol: 5.9182e-13 +- 5.3801e-13 [5.5067e-14 - 1.1653e-12] (24#)

Frozen and Alive Status

SingleDerivativeTester.java:156 executed in 0.01 seconds (0.000 gc):

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

LayerTests.java:425 executed in 0.00 seconds (0.000 gc):

    throwException(exceptions.addRef());

Results

classdetailsresult
com.simiacryptus.mindseye.test.unit.SingleDerivativeTesterToleranceStatistics{absoluteTol=1.5782e-13 +- 5.6238e-13 [0.0000e+00 - 2.3306e-12] (180#), relativeTol=5.9182e-13 +- 5.3801e-13 [5.5067e-14 - 1.1653e-12] (24#)}OK
 
  {
    "result": "OK",
    "performance": {
      "execution_time": "0.204",
      "gc_time": "0.112"
    },
    "created_on": 1586737003292,
    "file_name": "derivativeTest",
    "report": {
      "simpleName": "Basic",
      "canonicalName": "com.simiacryptus.mindseye.layers.java.ImgBandBiasLayerTest.Basic",
      "link": "https://github.com/SimiaCryptus/mindseye-java/tree/93db34cedee48c0202777a2b25deddf1dfaf5731/src/test/java/com/simiacryptus/mindseye/layers/java/ImgBandBiasLayerTest.java",
      "javaDoc": ""
    },
    "archive": "s3://code.simiacrypt.us/tests/com/simiacryptus/mindseye/layers/java/ImgBandBiasLayer/Basic/derivativeTest/202004131643",
    "id": "84d5a80e-6410-43de-97df-05808ef19425",
    "report_type": "Components",
    "display_name": "Derivative Validation",
    "target": {
      "simpleName": "ImgBandBiasLayer",
      "canonicalName": "com.simiacryptus.mindseye.layers.java.ImgBandBiasLayer",
      "link": "https://github.com/SimiaCryptus/mindseye-java/tree/93db34cedee48c0202777a2b25deddf1dfaf5731/src/main/java/com/simiacryptus/mindseye/layers/java/ImgBandBiasLayer.java",
      "javaDoc": ""
    }
  }