apunkt/GPy
1
0
Fork 0
mirror of https://github.com/SheffieldML/GPy.git synced 2026-04-25 21:06:23 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions
GPy/doc/tuto_interacting_with_models.rst
mzwiessele b9e897c50d [documentation] updated big parts of the doc
2014-09-07 15:42:03 +01:00

341 lines
15 KiB
ReStructuredText
Raw Blame History

.. _interacting_with_models:
*************************************
Interacting with models
*************************************
The GPy model class has a set of features which are
designed to make it simple to explore the parameter
space of the model. By default, the scipy optimisers
are used to fit GPy models (via model.optimize()),
for which we provide mechanisms for 'free' optimisation:
GPy can ensure that naturally positive parameters
(such as variances) remain positive. But these mechanisms
are much more powerful than simple reparameterisation,
as we shall see.
Along this tutorial we'll use a sparse GP regression model
as example. This example can be in ``GPy.examples.regression``.
All of the examples included in GPy return an instance
of a model class, and therefore they can be called in
the following way: ::
import numpy as np
import pylab as pb
pb.ion()
import GPy
m = GPy.examples.regression.sparse_GP_regression_1D()
Examining the model using print
===============================
To see the current state of the model parameters,
and the model's (marginal) likelihood just print the model ::
print m
The first thing displayed on the screen is the log-likelihood
value of the model with its current parameters. Below the
log-likelihood, a table with all the model's parameters
is shown. For each parameter, the table contains the name
of the parameter, the current value, and in case there are
defined: constraints, ties and prior distrbutions associated. ::
Name : sparse gp
Log-likelihood : 588.947189413
Number of Parameters : 8
Parameters:
sparse_gp. | Value | Constraint | Prior | Tied to
inducing inputs | (5, 1) | | |
rbf.variance | 1.91644016819 | +ve | |
rbf.lengthscale | 2.62103621347 | +ve | |
Gaussian_noise.variance | 0.00269870373421 | +ve | |
In this case the kernel parameters (``rbf.variance``,
``rbf.lengthscale``) as well as
the likelihood noise parameter (``Gaussian_noise.variance``), are constrained
to be positive, while the inducing inputs have no
constraints associated. Also there are no ties or prior defined.
You can also print all subparts of the model, by printing the
subcomponents individually::
print m.rbf
This will print the details of this particular parameter handle::
rbf. | Value | Constraint | Prior | Tied to
variance | 1.91644016819 | +ve | |
lengthscale | 2.62103621347 | +ve | |
When you want to get a closer look into
multivalue parameters, print them directly::
print m.inducing_inputs
Index | sparse_gp.inducing_inputs | Constraint | Prior | Tied to
[0 0] | 2.7189499 | | | N/A
[1 0] | 0.02006533 | | | N/A
[2 0] | -1.5299386 | | | N/A
[3 0] | -2.7001675 | | | N/A
[4 0] | 1.4654162 | | | N/A
Interacting with Parameters:
=======================
The preferred way of interacting with parameters is to act on the
parameter handle itself.
Interacting with parameter handles is simple. The names, printed by `print m`
are accessible interactively and programatically. For example try to
set kernels (`rbf`) `lengthscale` to `.2` and print the result::
m.rbf.lengthscale = .2
print m
You should see this::
Name : sparse gp
Log-likelihood : 588.947189413
Number of Parameters : 8
Parameters:
sparse_gp. | Value | Constraint | Prior | Tied to
inducing inputs | (5, 1) | | |
rbf.variance | 1.91644016819 | +ve | |
rbf.lengthscale | 0.2 | +ve | |
Gaussian_noise.variance | 0.00269870373421 | +ve | |
This will already have updated the model's inner state, so you can
plot it or see the changes in the posterior `m.posterior` of the model.
Regular expressions
----------------
The model's parameters can also be accessed through regular
expressions, by 'indexing' the model with a regular expression,
matching the parameter name. Through indexing by regular expression,
you can only retrieve leafs of the hierarchy, and you can retrieve the
values matched by calling `values()` on the returned object::
>>> print m['.*var']
Index | sparse_gp.rbf.variance | Constraint | Prior | Tied to
[0] | 2.1500132 | | | N/A
----- | sparse_gp.Gaussian_noise.variance | ---------- | ---------- | -------
[0] | 0.0024268215 | | | N/A
>>> print m['.*var'].values()
[ 2.1500132 0.00242682]
>>> print m['rbf']
Index | sparse_gp.rbf.variance | Constraint | Prior | Tied to
[0] | 2.1500132 | | | N/A
----- | sparse_gp.rbf.lengthscale | ---------- | ---------- | -------
[0] | 2.6782803 | | | N/A
There is access to setting parameters by regular expression,
as well. Here are a few examples of how to set parameters by regular expression::
>>> m['.*var'] = .1
>>> print m['.*var']
Index | sparse_gp.rbf.variance | Constraint | Prior | Tied to
[0] | 0.1 | | | N/A
----- | sparse_gp.Gaussian_noise.variance | ---------- | ---------- | -------
[0] | 0.1 | | | N/A
>>> m['.*var'] = [.1, .2]
>>> print m['.*var']
Index | sparse_gp.rbf.variance | Constraint | Prior | Tied to
[0] | 0.1 | | | N/A
----- | sparse_gp.Gaussian_noise.variance | ---------- | ---------- | -------
[0] | 0.2 | | | N/A
The fact that only leaf nodes can be accesses we can print all
parameters in a flattened view, by printing the regular expression
match of matching all objects::
>>> print m['']
Index | sparse_gp.inducing_inputs | Constraint | Prior | Tied to
[0 0] | -2.6716041 | | | N/A
[1 0] | -1.4665111 | | | N/A
[2 0] | -0.031010293 | | | N/A
[3 0] | 1.4563711 | | | N/A
[4 0] | 2.6803046 | | | N/A
----- | sparse_gp.rbf.variance | ---------- | ---------- | -------
[0] | 0.1 | | | N/A
----- | sparse_gp.rbf.lengthscale | ---------- | ---------- | -------
[0] | 2.6782803 | | | N/A
----- | sparse_gp.Gaussian_noise.variance | ---------- | ---------- | -------
[0] | 0.2 | | | N/A
Setting and fetching parameters `parameter_array`
------------------------------------------
Another way to interact with the model's parameters is through the
`parameter_array`. The Parameter array holds all the parameters of the
model in one place and is editable. It can be accessed through
indexing the model for example you can set all the parameters through
this mechanism::
>>> new_params = np.r_[[-4,-2,0,2,4], [.5,2], [.3]]
>>> print new_params
array([-4. , -2. , 0. , 2. , 4. , 0.5, 2. , 0.3])
>>> m[:] = new_params
>>> print m
Name : sparse gp
Log-likelihood : -147.561160209
Number of Parameters : 8
Parameters:
sparse_gp. | Value | Constraint | Prior | Tied to
inducing inputs | (5, 1) | | |
rbf.variance | 0.5 | +sq | |
rbf.lengthscale | 2.0 | +ve | |
Gaussian_noise.variance | 0.3 | +sq | |
Parameters themselves (leafs of the hierarchy) can be indexed and used
the same way as numpy arrays. First let us set a slice of the
`inducing_inputs`::
>>> m.inducing_inputs[2:, 0] = [1,3,5]
>>> print m.inducing_indputs
Index | sparse_gp.inducing_inputs | Constraint | Prior | Tied to
[0 0] | -4 | | | N/A
[1 0] | -2 | | | N/A
[2 0] | 1 | | | N/A
[3 0] | 3 | | | N/A
[4 0] | 5 | | | N/A
Or you use the parameters as normal numpy arrays for calculations::
>>> precision = 1./m.Gaussian_noise.variance
array([ 3.33333333])
Getting the model's log likelihood
=============================================
Appart form the printing the model, the marginal
log-likelihood can be obtained by using the function
``log_likelihood()``.::
>>> m.log_likelihood()
array([-152.83377316])
If you want to ensure the log likelihood as a float, call `float()`
around it::
>>> float(m.log_likelihood())
-152.83377316356177
Getting the model parameter's gradients
============================
The gradients of a model can shed light on understanding the
(possibly hard) optimization process. The gradients of each parameter
handle can be accessed through their `gradient` field.::
>>> print m.gradient
[ 5.51170031 9.71735112 -4.20282106 -3.45667035 -1.58828165
-2.11549358 12.40292787 -627.75467803]
>>> print m.rbf.gradient
[ -2.11549358 12.40292787]
>>> m.optimize()
>>> print m.gradient
[ -5.98046560e-04 -3.64576085e-04 1.98005930e-04 3.43381219e-04
-6.85685104e-04 -1.28800748e-05 1.08552429e-03 2.74058081e-01]
Adjusting the model's constraints
================================
When we initially call the example, it was optimized and hence the
log-likelihood gradients were close to zero. However, since
we have been changing the parameters, the gradients are far from zero now.
Next we are going to show how to optimize the model setting different
restrictions on the parameters.
Once a constraint has been set on a parameter, it is possible to remove
it with the command ``unconstrain()``, which can be called on any
parameter handle of the model. The methods `constrain()` and
`unconstrain()` return the indices which were actually unconstrained,
relative to the parameter handle the method was called on. This is
particularly handy for reporting which parameters where reconstrained,
when reconstraining a parameter, which was already constrained::
>>> m.rbf.variance.unconstrain()
array([0])
>>>m.unconstrain()
array([6, 7])
If you want to unconstrain only a specific constraint, you can pass it
as an argument of ``unconstrain(Transformation)`` (:py:class:`~GPy.constraints.Transformation`), or call
the respective method, such as ``unconstrain_fixed()`` (or
``unfix()``) to only unfix fixed parameters.::
>>> m.inducing_input[0].fix()
>>> m.unfix()
>>> m.rbf.constrain_positive()
>>> print m
Name : sparse gp
Log-likelihood : 620.741066698
Number of Parameters : 8
Parameters:
sparse_gp. | Value | Constraint | Prior | Tied to
inducing inputs | (5, 1) | | |
rbf.variance | 1.48329711218 | +ve | |
rbf.lengthscale | 2.5430947048 | +ve | |
Gaussian_noise.variance | 0.00229714444128 | | |
As you can see, ``unfix()`` only unfixed the inducing_input, and did
not change the positive constraint of the kernel.
The parameter handles come with default constraints, so you will
rarely be needing to adjust the constraints of a model. In the rare
cases of needing to adjust the constraints of a model, or in need of
fixing some parameters, you can do so with the functions
``constrain_{positive|negative|bounded|fixed}()``.::
m['.*var'].constrain_positive()
Available Constraints
==============
* :py:meth:`~GPy.constraints.Logexp`
* :py:meth:`~GPy.constraints.Exponent`
* :py:meth:`~GPy.constraints.Square`
* :py:meth:`~GPy.constraints.Logistic`
* :py:meth:`~GPy.constraints.LogexpNeg`
* :py:meth:`~GPy.constraints.NegativeExponent`
* :py:meth:`~GPy.constraints.NegativeLogexp`
Tying Parameters
============
Not yet implemented for GPy version 0.6.0
Optimizing the model
====================
Once we have finished defining the constraints,
we can now optimize the model with the function
``optimize``.::
m.Gaussian_noise.constrain_positive()
m.rbf.constrain_positive()
m.optimize()
By deafult, GPy uses the lbfgsb optimizer.
Some optional parameters may be discussed here.
* ``optimizer``: which optimizer to use, currently there are ``lbfgsb, fmin_tnc,
scg, simplex`` or any unique identifier uniquely identifying an
optimizer. Thus, you can say ``m.optimize('bfgs') for using the
``lbfgsb`` optimizer
* ``messages``: if the optimizer is verbose. Each optimizer has its
own way of printing, so do not be confused by differing messages of
different optimizers
* ``max_iters``: Maximum number of iterations to take. Some optimizers
see iterations as function calls, others as iterations of the
algorithm. Please be advised to look into ``scipy.optimize`` for
more instructions, if the number of iterations matter, so you can
give the right parameters to ``optimize()``
* ``gtol``: only for some optimizers. Will determine the convergence
criterion, as the tolerance of gradient to finish the optimization.
Further Reading
===============
All of the mechansiams for dealing
with parameters are baked right into GPy.core.model, from which all of
the classes in GPy.models inherrit. To learn how to construct your own
model, you might want to read :ref:`creating_new_models`. If you want
to learn how to create kernels, please refer to
:ref:`creating_new_kernels`
Reference in a new issue View git blame Copy permalink