evocov 0.1.0

GPlib extension to learn the kernel function.

EvoCov

GPlib extension to learn the kernel function.

Setup evocov

• Create and activate virtualenv (for python2) or venv (for python3)

  # for python3
  python3 -m venv .env
  # or for python2
  python2 -m virtualenv .env

  source .env/bin/activate

• Upgrade pip

  python -m pip install --upgrade pip

• Install EvoCov package

  python -m pip install evocov

Use EvoCov

• Import EvoCov and GPlib to use it in your python script.

  import gplib
  import evocov

• Configure the fitting method.

  lml = gplib.me.LML()
  bic = gplib.me.BIC()

  fitting_method = evocov.fit.EvoCov(
      obj_fun=bic.fold_measure,
      max_fun_call=25000,
      nested_fit_method=gplib.fit.MultiStart(
          obj_fun=lml.fold_measure,
          max_fun_call=250,
          nested_fit_method=gplib.fit.LocalSearch(
              obj_fun=lml.fold_measure,
              method="Powell",
              max_fun_call=100
          )
      )
  )

• Initialize the GP with None covariance function.

  gp = gplib.GP(
      mean_function=gplib.mea.Fixed(),
      covariance_function=fitting_method.get_random_kernel()
  )

• Generate some random data.

  import numpy as np
  data = {
    'X': np.arange(3, 8, 1.)[:, None],
    'Y': np.random.uniform(0, 2, 5)[:, None]
  }

• Fit the kernel and the hyperparameters to the training set.

  validation = gplib.dm.Full()

  log = fitting_method.fit(gp, validation.get_folds(
      data
  ))

• Finally plot the posterior GP.

  posterior_gp = gp.get_posterior(data)
  gplib.plot.gp_1d(posterior_gp, data, n_samples=10)

• There are more examples in examples/ directory. Check them out!

Develop EvoCov

• Update API documentation

  source ./.env/bin/activate
  pip install Sphinx
  cd docs/
  sphinx-apidoc -f -o ./ ../evocov