modifying_surrogate_model.rst.txt 1.35 KB
Newer Older
1
2
.. _modify_gpr:

3
4
5
6
=========================
Modifying surrogate model
=========================

7
8
This tutorial extends the previous one for
:ref:`Cu15 clusters <cu_cluster_search>`. It is
9
10
11
12
therefore recomended that you do that one before the present one.

In the avove mentioned tutorial GOFEE was initialized with the following
arguments::
13

14
    from gofee import GOFEE
15
16
17
    search = GOFEE(calc=calc,
                   startgenerator=sg,
                   candidate_generator=candidate_generator,
18
                   max_steps=60,
19
20
21
22
23
24
25
                   population_size=5)

however GOFEE takes a number of other arguments, including a
Gaussian Process regression (GPR) model, which is actively learned
during the search and used for cheap optimization of new candidates.

One can for example apply a GPR model with another degree of regularization
26
27
in the search. This is controlled by the ``noise`` parameter of the ``kernel``,
passed to the GPR model. The modification can be achieved by::
28

29
30
    from gofee.surrogate import GPR
    from gofee.surrogate.kernel import DoubleGaussKernel
31

32
    kernel = DoubleGaussKernel(noise=1e-6)
33
34
35
36
37
38
    gpr = GPR(kernel=kernel)

    search = GOFEE(calc=calc,
                   gpr=gpr,
                   startgenerator=sg,
                   candidate_generator=candidate_generator,
39
                   max_steps=60,
40
41
                   population_size=5)