public void CalculateNonZero_Average_With_No_Data_Returns_Zero()
{
var samples = new WorkoutSamples(0);
var vector = new WorkoutSampleVector(3, WorkoutSampleDataType.Power);
IAthlete athlete = new Athlete();
var workoutCalculator = new WorkoutSamplesCalculator(samples, athlete);
var average = workoutCalculator.CalculateNonZeroVectorAverage(vector);
Assert.Equal(0, average);
}
public void CalculatCadenceClassificatoion_Under_2_Samples_returns_()
{
var samples = new WorkoutSamples(0);
var vector = new WorkoutSampleVector(1, WorkoutSampleDataType.Cadence);
vector.AddPoint(1, 2);
samples.CadenceVector = vector;
IAthlete athlete = new Athlete();
var workoutCalculator = new WorkoutSamplesCalculator(samples, athlete);
var classification = workoutCalculator.ClassifyWorkoutCadenceRanges();
Assert.IsType <List <ICadenceRange> >(classification);
}
public void CalculateNonZero_Average_Returns_Average()
{
var samples = new WorkoutSamples(0);
var vector = new WorkoutSampleVector(3, WorkoutSampleDataType.Power);
vector.AddPoint(1, 2);
vector.AddPoint(2, 4);
vector.AddPoint(3, 0);
IAthlete athlete = new Athlete();
var workoutCalculator = new WorkoutSamplesCalculator(samples, athlete);
var average = workoutCalculator.CalculateNonZeroVectorAverage(vector);
Assert.Equal(3, average);
}
public void CalculateNormalizedPower_With_Under_30_Samples_Returns_Average()
{
//Initialise();
var samples = new WorkoutSamples(0);
var vector = new WorkoutSampleVector(2, WorkoutSampleDataType.Power);
vector.AddPoint(1, 2);
vector.AddPoint(2, 4);
IAthlete athlete = new Athlete();
var workoutCalculator = new WorkoutSamplesCalculator(samples, athlete);
var average = workoutCalculator.CalculateVectorNormalizedAverage(vector);
Assert.Equal(3, average);
}
public double CalculateVectorNormalizedAverage(WorkoutSampleVector vector)
{
if (!vector.HasData)
{
return(0);
}
if (vector.NumberOfSamples < 30)
{
return(CalculateVectorAverage(vector));
}
var movingAverageBuffer = new SimpleMovingAverage(30);
double movingAverage = 0;
var movingAverageSamples = 1;
movingAverageBuffer.Add(vector.Vector[0].dataPoint);
for (var i = 1; i < vector.NumberOfSamples; i++)
{
if (vector.Vector[i].dataPoint < 0)
{
continue;
}
var timeDiff = vector.Vector[i].timeOffsetSeconds - vector.Vector[i - 1].timeOffsetSeconds;
for (var j = 0; j < timeDiff; j++)
//Cannot multiply value across time in the sample as we need a 30s moving avg.
{
movingAverageBuffer.Add(vector.Vector[i].dataPoint);
if (movingAverageBuffer.NumberOfSamples >= 30)
{
movingAverage += Math.Pow(Math.Round(movingAverageBuffer.Average()), 4);
movingAverageSamples++;
}
}
}
var average = movingAverage / movingAverageSamples;
average = Math.Round(Math.Pow(average, 0.25));
return(average);
}
internal double CalculateVectorAverage(WorkoutSampleVector vector)
{
if (!vector.HasData)
{
return(0);
}
var weightedCumulative = vector.Vector[0].dataPoint;
for (var i = 1; i < vector.NumberOfSamples; i++)
{
if (vector.Vector[i].dataPoint <= 0)
{
continue;
}
var timeDiff = vector.Vector[i].timeOffsetSeconds - vector.Vector[i - 1].timeOffsetSeconds;
weightedCumulative = weightedCumulative + timeDiff * vector.Vector[i].dataPoint;
}
var divisor = vector.Vector[vector.NumberOfSamples - 1].timeOffsetSeconds;
var average = weightedCumulative / divisor;
return(Math.Round(average));
}
internal IEnumerable <IRange> ClassifyWorkoutCadenceType(IEnumerable <IRange> ranges, WorkoutSampleVector vector, double referenceValue)
{
if (vector.NumberOfSamples < 2)
{
return(ranges);
}
var rangeList = ranges as IList <IRange> ?? ranges.ToList();
ClassifyCadenceDataPoint(vector.Vector[0].dataPoint, rangeList, referenceValue);
for (var i = 1; i < vector.NumberOfSamples; i++)
{
if (!(vector.Vector[i].dataPoint >= 0))
{
continue;
}
var timeDiff = vector.Vector[i].timeOffsetSeconds -
vector.Vector[i - 1].timeOffsetSeconds;
for (var j = 0; j < timeDiff; j++)
{
ClassifyCadenceDataPoint(vector.Vector[i].dataPoint, rangeList, referenceValue);
}
}
var totalSamples = rangeList.Sum(range => range.QuanityOfSamples);
foreach (var range in rangeList)
{
var percent = range.QuanityOfSamples / (double)totalSamples * 100;
range.PercentOfTotal = Math.Round(percent);
}
return(rangeList);
}