mvpa2.clfs.statsΒΆ
Estimator for classifier error distributions.
Functions
auto_null_dist(dist) |
Cheater for human beings – wraps dist if needed with some |
is_datasetlike(obj) |
Check if an object looks like a Dataset. |
kstest(rvs, cdf[, args, N, alternative, mode]) |
Perform the Kolmogorov-Smirnov test for goodness of fit. |
match_distribution(data[, nsamples, loc, ...]) |
Determine best matching distribution. |
nanmean(x[, axis]) |
Compute the mean over the given axis ignoring NaNs. |
plot_distribution_matches(data, matches[, ...]) |
Plot best matching distributions |
Classes
AdaptiveNormal(dist, \*\*kwargs) |
Adaptive Normal Distribution: params are (0, sqrt(1/nfeatures)) |
AdaptiveNullDist(dist, \*\*kwargs) |
Adaptive distribution which adjusts parameters according to the data |
AdaptiveRDist(dist, \*\*kwargs) |
Adaptive rdist: params are (nfeatures-1, 0, 1) |
AttributePermutator(attr[, count, limit, ...]) |
Node to permute one a more attributes in a dataset. |
ClassWithCollections([descr]) |
Base class for objects which contain any known collection |
ConditionalAttribute([enabled]) |
Simple container intended to conditionally store the value |
Dataset(samples[, sa, fa, a]) |
Generic storage class for datasets with multiple attributes. |
FixedNullDist(dist, \*\*kwargs) |
Proxy/Adaptor class for SciPy distributions. |
MCNullDist(permutator[, dist_class, measure]) |
Null-hypothesis distribution is estimated from randomly permuted data labels. |
Nonparametric(dist_samples[, correction]) |
Non-parametric 1d distribution – derives cdf based on stored values. |
NullDist([tail]) |
Base class for null-hypothesis testing. |
rv_semifrozen(dist[, loc, scale, args]) |
Helper proxy-class to fit distribution when some parameters are known |
