stats
Implements statistical significance testing for the output of two systems by randomization.
Provides utilities for descriptive statistics, like the mean and variance.
A breeze.generic.UFunc for counting bins.
A breeze.generic.UFunc for counting bins.
If passed a traversable object full of Int's, provided those ints are larger than 0, it will return an array of the bin counts. E.g.: bincount(DenseVector[Int](0,1,2,3,1,3,3,3)) == DenseVector[Int](1,2,1,4)
One can also call this on two dense vectors - the second will be treated as an array of weights. E.g.: val x = DenseVector[Int](0,1,2,3,1) val weights = DenseVector[Double](1.0,2.0,1.0,7.0,1.0) result is bincount(x, weights) == DenseVector[Double](1.0,3.0,1,7.0)
A breeze.generic.UFunc for digitizing arrays.
A breeze.generic.UFunc for digitizing arrays.
Each element in the bins array is assumed to be the *right* endpoint of a given bin. For instance, bins=[1,3,5] represents a bin from (-infty,1], (1,3], (3,5] and (5,\infty). The result returned is the index of the bin of the inputs.
E.g., digitize([-3, 0.5, 1, 1.5, 4], [0,1,2]) = [0, 1, 1, 2, 3]
This package contains hypothesis tests.
A breeze.generic.UFunc for computing the mean of objects
A breeze.generic.UFunc for computing the mean and sample variance of objects.
A breeze.generic.UFunc for computing the mean and sample variance of objects. This uses an efficient, numerically stable, one pass algorithm for computing both the mean and the variance.
A breeze.generic.UFunc for computing the median of objects
Computes the *sample* standard deviation by calling variance and then sqrt'ing.
Computes the *sample* standard deviation by calling variance and then sqrt'ing. Note that this is different from Excel, numpy, etc.
Call stddev.population if you want that.
A breeze.generic.UFunc for computing the *sample* variance of objects.
A breeze.generic.UFunc for computing the *sample* variance of objects. The method just calls meanAndVariance and returns the second result.
Call variance.population if you want population variance
Implements statistical significance testing for the output of two systems by randomization. This system assumes they're on the same dataset, which changes the procedure. Follows Teh, 2000 More accurate tests for the statistical significance of result differences.
Labels must have .equals.