diff options
| -rw-r--r-- | networkQuality.go | 4 | ||||
| -rw-r--r-- | qualityattenuation/qualityattenuation.go | 112 | ||||
| -rw-r--r-- | qualityattenuation/qualityattenuation_test.go | 44 |
3 files changed, 82 insertions, 78 deletions
diff --git a/networkQuality.go b/networkQuality.go index 676d98d..10ee8d3 100644 --- a/networkQuality.go +++ b/networkQuality.go @@ -550,7 +550,9 @@ timeout: } } else if probeMeasurement.Type == probe.SelfDown || probeMeasurement.Type == probe.SelfUp { selfRtts.AddElement(probeMeasurement.Duration.Seconds()) - selfRttsQualityAttenuation.AddSample(probeMeasurement.Duration.Seconds()) + if *printQualityAttenuation { + selfRttsQualityAttenuation.AddSample(probeMeasurement.Duration.Seconds()) + } } if probeMeasurement.Type == probe.Foreign { diff --git a/qualityattenuation/qualityattenuation.go b/qualityattenuation/qualityattenuation.go index 980253f..279959e 100644 --- a/qualityattenuation/qualityattenuation.go +++ b/qualityattenuation/qualityattenuation.go @@ -3,79 +3,81 @@ package qualityattenuation import ( + "fmt" "math" "github.com/influxdata/tdigest" ) type SimpleQualityAttenuation struct { - EmpiricalDistribution *tdigest.TDigest - Offset float64 - Offset_sum float64 - Offset_sum_of_squares float64 - Number_of_samples int64 - Number_of_losses int64 - Latency_eq_loss_threshold float64 - Minimum_latency float64 - Maximum_latency float64 + empiricalDistribution *tdigest.TDigest + offset float64 + offsetSum float64 + offsetSumOfSquares float64 + numberOfSamples int64 + numberOfLosses int64 + latencyEqLossThreshold float64 + minimumLatency float64 + maximumLatency float64 } func NewSimpleQualityAttenuation() *SimpleQualityAttenuation { return &SimpleQualityAttenuation{ - EmpiricalDistribution: tdigest.NewWithCompression(50), - Offset: 0.1, - Offset_sum: 0.0, - Offset_sum_of_squares: 0.0, - Number_of_samples: 0, - Number_of_losses: 0, - Latency_eq_loss_threshold: 15.0, // Count latency greater than this value as a loss. - Minimum_latency: 0.0, - Maximum_latency: 0.0, + empiricalDistribution: tdigest.NewWithCompression(50), + offset: 0.1, + offsetSum: 0.0, + offsetSumOfSquares: 0.0, + numberOfSamples: 0, + numberOfLosses: 0, + latencyEqLossThreshold: 15.0, // Count latency greater than this value as a loss. + minimumLatency: 0.0, + maximumLatency: 0.0, } } -func (qa *SimpleQualityAttenuation) AddSample(sample float64) { +func (qa *SimpleQualityAttenuation) AddSample(sample float64) error { if sample <= 0.0 { // Ignore zero or negative samples because they cannot be valid. // TODO: This should raise a warning and/or trigger error handling. - return + return fmt.Errorf("sample is zero or negative") } - qa.Number_of_samples++ - if sample > qa.Latency_eq_loss_threshold { - qa.Number_of_losses++ - return + qa.numberOfSamples++ + if sample > qa.latencyEqLossThreshold { + qa.numberOfLosses++ + return nil } else { - if qa.Minimum_latency == 0.0 || sample < qa.Minimum_latency { - qa.Minimum_latency = sample + if qa.minimumLatency == 0.0 || sample < qa.minimumLatency { + qa.minimumLatency = sample } - if qa.Maximum_latency == 0.0 || sample > qa.Maximum_latency { - qa.Maximum_latency = sample + if qa.maximumLatency == 0.0 || sample > qa.maximumLatency { + qa.maximumLatency = sample } - qa.EmpiricalDistribution.Add(sample, 1) - qa.Offset_sum += sample - qa.Offset - qa.Offset_sum_of_squares += (sample - qa.Offset) * (sample - qa.Offset) + qa.empiricalDistribution.Add(sample, 1) + qa.offsetSum += sample - qa.offset + qa.offsetSumOfSquares += (sample - qa.offset) * (sample - qa.offset) } + return nil } func (qa *SimpleQualityAttenuation) GetNumberOfLosses() int64 { - return qa.Number_of_losses + return qa.numberOfLosses } func (qa *SimpleQualityAttenuation) GetNumberOfSamples() int64 { - return qa.Number_of_samples + return qa.numberOfSamples } func (qa *SimpleQualityAttenuation) GetPercentile(percentile float64) float64 { - return qa.EmpiricalDistribution.Quantile(percentile / 100) + return qa.empiricalDistribution.Quantile(percentile / 100) } func (qa *SimpleQualityAttenuation) GetAverage() float64 { - return qa.Offset_sum/float64(qa.Number_of_samples-qa.Number_of_losses) + qa.Offset + return qa.offsetSum/float64(qa.numberOfSamples-qa.numberOfLosses) + qa.offset } func (qa *SimpleQualityAttenuation) GetVariance() float64 { - number_of_latency_samples := float64(qa.Number_of_samples) - float64(qa.Number_of_losses) - return (qa.Offset_sum_of_squares - (qa.Offset_sum * qa.Offset_sum / number_of_latency_samples)) / (number_of_latency_samples) + number_of_latency_samples := float64(qa.numberOfSamples) - float64(qa.numberOfLosses) + return (qa.offsetSumOfSquares - (qa.offsetSum * qa.offsetSum / number_of_latency_samples)) / (number_of_latency_samples - 1) } func (qa *SimpleQualityAttenuation) GetStandardDeviation() float64 { @@ -83,11 +85,11 @@ func (qa *SimpleQualityAttenuation) GetStandardDeviation() float64 { } func (qa *SimpleQualityAttenuation) GetMinimum() float64 { - return qa.Minimum_latency + return qa.minimumLatency } func (qa *SimpleQualityAttenuation) GetMaximum() float64 { - return qa.Maximum_latency + return qa.maximumLatency } func (qa *SimpleQualityAttenuation) GetMedian() float64 { @@ -95,7 +97,7 @@ func (qa *SimpleQualityAttenuation) GetMedian() float64 { } func (qa *SimpleQualityAttenuation) GetLossPercentage() float64 { - return 100 * float64(qa.Number_of_losses) / float64(qa.Number_of_samples) + return 100 * float64(qa.numberOfLosses) / float64(qa.numberOfSamples) } func (qa *SimpleQualityAttenuation) GetRPM() float64 { @@ -109,26 +111,26 @@ func (qa *SimpleQualityAttenuation) GetPDV(percentile float64) float64 { // Merge two quality attenuation values. This operation assumes the two samples have the same offset and latency_eq_loss_threshold, and // will return an error if they do not. // It also assumes that the two quality attenuation values are measurements of the same thing (path, outcome, etc.). -func (qa *SimpleQualityAttenuation) Merge(other *SimpleQualityAttenuation) { +func (qa *SimpleQualityAttenuation) Merge(other *SimpleQualityAttenuation) error { // Check that offsets are the same - if qa.Offset != other.Offset || - qa.Latency_eq_loss_threshold != other.Latency_eq_loss_threshold { - //"Cannot merge quality attenuation values with different offset or latency_eq_loss_threshold" - + if qa.offset != other.offset || + qa.latencyEqLossThreshold != other.latencyEqLossThreshold { + return fmt.Errorf("merge quality attenuation values with different offset or latency_eq_loss_threshold") } - for _, centroid := range other.EmpiricalDistribution.Centroids() { + for _, centroid := range other.empiricalDistribution.Centroids() { mean := centroid.Mean weight := centroid.Weight - qa.EmpiricalDistribution.Add(mean, weight) + qa.empiricalDistribution.Add(mean, weight) } - qa.Offset_sum += other.Offset_sum - qa.Offset_sum_of_squares += other.Offset_sum_of_squares - qa.Number_of_samples += other.Number_of_samples - qa.Number_of_losses += other.Number_of_losses - if other.Minimum_latency < qa.Minimum_latency { - qa.Minimum_latency = other.Minimum_latency + qa.offsetSum += other.offsetSum + qa.offsetSumOfSquares += other.offsetSumOfSquares + qa.numberOfSamples += other.numberOfSamples + qa.numberOfLosses += other.numberOfLosses + if other.minimumLatency < qa.minimumLatency { + qa.minimumLatency = other.minimumLatency } - if other.Maximum_latency > qa.Maximum_latency { - qa.Maximum_latency = other.Maximum_latency + if other.maximumLatency > qa.maximumLatency { + qa.maximumLatency = other.maximumLatency } + return nil } diff --git a/qualityattenuation/qualityattenuation_test.go b/qualityattenuation/qualityattenuation_test.go index 975cbf3..4746364 100644 --- a/qualityattenuation/qualityattenuation_test.go +++ b/qualityattenuation/qualityattenuation_test.go @@ -11,23 +11,23 @@ func TestBasicSimpleQualityAttenuation(t *testing.T) { qa.AddSample(1.0) qa.AddSample(2.0) qa.AddSample(3.0) - assert.Equal(t, qa.Number_of_samples, int64(3)) - assert.Equal(t, qa.Number_of_losses, int64(0)) - assert.InEpsilon(t, 1.0, qa.Minimum_latency, 0.000001) - assert.InEpsilon(t, 3.0, qa.Maximum_latency, 0.000001) - assert.InEpsilon(t, 5.7, qa.Offset_sum, 0.000001) - assert.InEpsilon(t, 12.83, qa.Offset_sum_of_squares, 0.000001) - assert.InEpsilon(t, 1.0, qa.EmpiricalDistribution.Quantile(0.1), 0.000001) - assert.InEpsilon(t, 2.0, qa.EmpiricalDistribution.Quantile(0.5), 0.000001) - assert.InEpsilon(t, 3.0, qa.EmpiricalDistribution.Quantile(0.9), 0.000001) + assert.Equal(t, qa.numberOfSamples, int64(3)) + assert.Equal(t, qa.numberOfLosses, int64(0)) + assert.InEpsilon(t, 1.0, qa.minimumLatency, 0.000001) + assert.InEpsilon(t, 3.0, qa.maximumLatency, 0.000001) + assert.InEpsilon(t, 5.7, qa.offsetSum, 0.000001) + assert.InEpsilon(t, 12.83, qa.offsetSumOfSquares, 0.000001) + assert.InEpsilon(t, 1.0, qa.empiricalDistribution.Quantile(0.1), 0.000001) + assert.InEpsilon(t, 2.0, qa.empiricalDistribution.Quantile(0.5), 0.000001) + assert.InEpsilon(t, 3.0, qa.empiricalDistribution.Quantile(0.9), 0.000001) //Test the get functions assert.Equal(t, qa.GetNumberOfSamples(), int64(3)) assert.Equal(t, qa.GetNumberOfLosses(), int64(0)) assert.InEpsilon(t, 1.0, qa.GetMinimum(), 0.000001) assert.InEpsilon(t, 3.0, qa.GetMaximum(), 0.000001) assert.InEpsilon(t, 2.0, qa.GetAverage(), 0.000001) - assert.InEpsilon(t, 0.666667, qa.GetVariance(), 0.000001) - assert.InEpsilon(t, 0.816496, qa.GetStandardDeviation(), 0.000001) + assert.InEpsilon(t, 1.000000, qa.GetVariance(), 0.000001) + assert.InEpsilon(t, 1.000000, qa.GetStandardDeviation(), 0.000001) assert.InEpsilon(t, 2.0, qa.GetMedian(), 0.000001) assert.InEpsilon(t, 1.0, qa.GetLossPercentage()+1.000000000, 0.000001) assert.InEpsilon(t, 30, qa.GetRPM(), 0.000001) @@ -42,23 +42,23 @@ func TestManySamples(t *testing.T) { for i := 1; i < 160000; i++ { qa.AddSample(float64(i) / 10000.0) //Linear ramp from 0.0001 to 16.0 } - assert.Equal(t, qa.Number_of_samples, int64(160000-1)) - assert.Equal(t, qa.Number_of_losses, int64(10000-1)) //Samples from 15.0001 to 16.0 are lost - assert.InEpsilon(t, 0.0001, qa.Minimum_latency, 0.000001) - assert.InEpsilon(t, 15.0000, qa.Maximum_latency, 0.000001) - assert.InEpsilon(t, 1110007.5, qa.Offset_sum, 0.000001) - assert.InEpsilon(t, 11026611.00024998, qa.Offset_sum_of_squares, 0.000001) - assert.InEpsilon(t, 1.50005, qa.EmpiricalDistribution.Quantile(0.1), 0.000001) - assert.InEpsilon(t, 7.500049, qa.EmpiricalDistribution.Quantile(0.5), 0.000001) - assert.InEpsilon(t, 13.50005, qa.EmpiricalDistribution.Quantile(0.9), 0.000001) + assert.Equal(t, qa.numberOfSamples, int64(160000-1)) + assert.Equal(t, qa.numberOfLosses, int64(10000-1)) //Samples from 15.0001 to 16.0 are lost + assert.InEpsilon(t, 0.0001, qa.minimumLatency, 0.000001) + assert.InEpsilon(t, 15.0000, qa.maximumLatency, 0.000001) + assert.InEpsilon(t, 1110007.5, qa.offsetSum, 0.000001) + assert.InEpsilon(t, 11026611.00024998, qa.offsetSumOfSquares, 0.000001) + assert.InEpsilon(t, 1.50005, qa.empiricalDistribution.Quantile(0.1), 0.000001) + assert.InEpsilon(t, 7.500049, qa.empiricalDistribution.Quantile(0.5), 0.000001) + assert.InEpsilon(t, 13.50005, qa.empiricalDistribution.Quantile(0.9), 0.000001) //Test the get functions assert.Equal(t, qa.GetNumberOfSamples(), int64(160000-1)) assert.Equal(t, qa.GetNumberOfLosses(), int64(10000-1)) assert.InEpsilon(t, 0.0001, qa.GetMinimum(), 0.000001) assert.InEpsilon(t, 15.0000, qa.GetMaximum(), 0.000001) assert.InEpsilon(t, 7.500049, qa.GetAverage(), 0.000001) - assert.InEpsilon(t, 18.749750, qa.GetVariance(), 0.001) - assert.InEpsilon(t, 4.330098, qa.GetStandardDeviation(), 0.001) + assert.InEpsilon(t, 18.750120, qa.GetVariance(), 0.000001) + assert.InEpsilon(t, 4.330141, qa.GetStandardDeviation(), 0.000001) assert.InEpsilon(t, 7.500049, qa.GetMedian(), 0.000001) assert.InEpsilon(t, 6.249414, qa.GetLossPercentage(), 0.000001) assert.InEpsilon(t, 7.999947, qa.GetRPM(), 0.000001) |