Skip to main content

Functional ANOVA using Gaussian Process priors.

Project description

# Gaussian Process Functional ANOVA

Implementation of a functional ANOVA (FANOVA) model, based partly on the model in

[Bayesian functional ANOVA modeling using Gaussian process prior distributions](http://projecteuclid.org/euclid.ba/1340369795). To implement a FANOVA model, an underlying general framework is defined for modeling functional observations:

$$ Y(t) = X beta(t),$$

where $$ Y(t) = [y_1(t),dots,y_m(t)]^T, $$ $$beta(t) = [beta_1(t),dots,beta_f(t)]^T,$$ $$ X: m times f$$ for a given time $t$. The design matrix $X$ defines the relation between the functions $beta$ and observations $y$. In general, the rank of $X$ should match the number of functions $f$. The FANOVA model can then be described by a specific form of $X$ such that

$$ y_{i,j}(t) = mu(t) + alpha_i(t) + beta_j(t) + alphabeta_{i,j}(t). $$

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

gpfanova-0.1.14.tar.gz (106.9 kB view hashes)

Uploaded Source

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page