From b235305bab20c6766364872ef9aed02b2eb787ee Mon Sep 17 00:00:00 2001
From: Manuel Guenther <manuel.guenther@idiap.ch>
Date: Thu, 5 Mar 2015 19:19:47 +0100
Subject: [PATCH] Updated MAP_GMMTrainer and removed the intermediate Python
class.
---
bob/learn/em/MAP_gmm_trainer.cpp | 109 ++++++++++++++--------------
bob/learn/em/__MAP_gmm_trainer__.py | 48 ------------
bob/learn/em/__init__.py | 2 -
bob/learn/em/test/test_em.py | 22 +++---
4 files changed, 67 insertions(+), 114 deletions(-)
delete mode 100644 bob/learn/em/__MAP_gmm_trainer__.py
diff --git a/bob/learn/em/MAP_gmm_trainer.cpp b/bob/learn/em/MAP_gmm_trainer.cpp
index 9868577..d62fb6f 100644
--- a/bob/learn/em/MAP_gmm_trainer.cpp
+++ b/bob/learn/em/MAP_gmm_trainer.cpp
@@ -17,28 +17,28 @@ static inline bool f(PyObject* o){return o != 0 && PyObject_IsTrue(o) > 0;} /*
static auto MAP_GMMTrainer_doc = bob::extension::ClassDoc(
BOB_EXT_MODULE_PREFIX ".MAP_GMMTrainer",
- "This class implements the maximum a posteriori M-step of the expectation-maximisation algorithm for a GMM Machine. The prior parameters are encoded in the form of a GMM (e.g. a universal background model). The EM algorithm thus performs GMM adaptation."
+ "This class implements the maximum a posteriori M-step of the expectation-maximization algorithm for a GMM Machine. The prior parameters are encoded in the form of a GMM (e.g. a universal background model). The EM algorithm thus performs GMM adaptation."
).add_constructor(
bob::extension::FunctionDoc(
"__init__",
"Creates a MAP_GMMTrainer",
- "",
+ "Additionally to the copy constructor, there are two different ways to call this constructor, one using the ``relevance_factor`` and one using the ``alpha``, both which have the same signature. "
+ "Hence, the only way to differentiate the two functions is by using keyword arguments.",
true
)
- .add_prototype("prior_gmm,relevance_factor, update_means, [update_variances], [update_weights], [mean_var_update_responsibilities_threshold]","")
- .add_prototype("prior_gmm,alpha, update_means, [update_variances], [update_weights], [mean_var_update_responsibilities_threshold]","")
+ .add_prototype("prior_gmm, relevance_factor, [update_means], [update_variances], [update_weights], [mean_var_update_responsibilities_threshold]","")
+ .add_prototype("prior_gmm, alpha, [update_means], [update_variances], [update_weights], [mean_var_update_responsibilities_threshold]","")
.add_prototype("other","")
- .add_parameter("prior_gmm", ":py:class:`bob.learn.em.GMMMachine`", "The prior GMM to be adapted (Universal Backgroud Model UBM).")
- .add_parameter("reynolds_adaptation", "bool", "Will use the Reynolds adaptation procedure? See Eq (14) from [Reynolds2000]_")
- .add_parameter("relevance_factor", "double", "If set the reynolds_adaptation parameters, will apply the Reynolds Adaptation procedure. See Eq (14) from [Reynolds2000]_")
- .add_parameter("alpha", "double", "Set directly the alpha parameter (Eq (14) from [Reynolds2000]_), ignoring zeroth order statistics as a weighting factor.")
+ .add_parameter("prior_gmm", ":py:class:`bob.learn.em.GMMMachine`", "The prior GMM to be adapted (Universal Background Model UBM).")
+ .add_parameter("relevance_factor", "float", "If set the Reynolds Adaptation procedure will be applied. See Eq (14) from [Reynolds2000]_")
+ .add_parameter("alpha", "float", "Set directly the alpha parameter (Eq (14) from [Reynolds2000]_), ignoring zeroth order statistics as a weighting factor.")
- .add_parameter("update_means", "bool", "Update means on each iteration")
- .add_parameter("update_variances", "bool", "Update variances on each iteration")
- .add_parameter("update_weights", "bool", "Update weights on each iteration")
- .add_parameter("mean_var_update_responsibilities_threshold", "float", "Threshold over the responsibilities of the Gaussians Equations 9.24, 9.25 of Bishop, `Pattern recognition and machine learning`, 2006 require a division by the responsibilities, which might be equal to zero because of numerical issue. This threshold is used to avoid such divisions.")
+ .add_parameter("update_means", "bool", "[Default: ``True``] Update means on each iteration")
+ .add_parameter("update_variances", "bool", "[Default: ``True``] Update variances on each iteration")
+ .add_parameter("update_weights", "bool", "[Default: ``True``] Update weights on each iteration")
+ .add_parameter("mean_var_update_responsibilities_threshold", "float", "[Default: min_float] Threshold over the responsibilities of the Gaussians Equations 9.24, 9.25 of Bishop, `Pattern recognition and machine learning`, 2006 require a division by the responsibilities, which might be equal to zero because of numerical issue. This threshold is used to avoid such divisions.")
.add_parameter("other", ":py:class:`bob.learn.em.MAP_GMMTrainer`", "A MAP_GMMTrainer object to be copied.")
);
@@ -62,7 +62,7 @@ static int PyBobLearnEMMAPGMMTrainer_init_base_trainer(PyBobLearnEMMAPGMMTrainer
char** kwlist1 = MAP_GMMTrainer_doc.kwlist(0);
char** kwlist2 = MAP_GMMTrainer_doc.kwlist(1);
-
+
PyBobLearnEMGMMMachineObject* gmm_machine;
bool reynolds_adaptation = false;
double alpha = 0.5;
@@ -78,27 +78,31 @@ static int PyBobLearnEMMAPGMMTrainer_init_base_trainer(PyBobLearnEMMAPGMMTrainer
PyObject* keyword_alpha = Py_BuildValue("s", kwlist2[1]);
auto keyword_relevance_factor_ = make_safe(keyword_relevance_factor);
- auto keyword_alpha_ = make_safe(keyword_alpha);
-
- //Here we have to select which keyword argument to read
- if (kwargs && PyDict_Contains(kwargs, keyword_relevance_factor) && (PyArg_ParseTupleAndKeywords(args, kwargs, "O!dO!|O!O!d", kwlist1,
- &PyBobLearnEMGMMMachine_Type, &gmm_machine,
- &aux,
- &PyBool_Type, &update_means,
- &PyBool_Type, &update_variances,
- &PyBool_Type, &update_weights,
- &mean_var_update_responsibilities_threshold)))
- reynolds_adaptation = true;
- else if (kwargs && PyDict_Contains(kwargs, keyword_alpha) && (PyArg_ParseTupleAndKeywords(args, kwargs, "O!dO!|O!O!d", kwlist2,
- &PyBobLearnEMGMMMachine_Type, &gmm_machine,
- &aux,
- &PyBool_Type, &update_means,
- &PyBool_Type, &update_variances,
- &PyBool_Type, &update_weights,
- &mean_var_update_responsibilities_threshold)))
+ auto keyword_alpha_ = make_safe(keyword_alpha);
+
+ //Here we have to select which keyword argument to read
+ if (kwargs && PyDict_Contains(kwargs, keyword_relevance_factor)){
+ if (!PyArg_ParseTupleAndKeywords(args, kwargs, "O!d|O!O!O!d", kwlist1,
+ &PyBobLearnEMGMMMachine_Type, &gmm_machine,
+ &aux,
+ &PyBool_Type, &update_means,
+ &PyBool_Type, &update_variances,
+ &PyBool_Type, &update_weights,
+ &mean_var_update_responsibilities_threshold))
+ return -1;
+ reynolds_adaptation = true;
+ } else if (kwargs && PyDict_Contains(kwargs, keyword_alpha)){
+ if (!PyArg_ParseTupleAndKeywords(args, kwargs, "O!d|O!O!O!d", kwlist2,
+ &PyBobLearnEMGMMMachine_Type, &gmm_machine,
+ &aux,
+ &PyBool_Type, &update_means,
+ &PyBool_Type, &update_variances,
+ &PyBool_Type, &update_weights,
+ &mean_var_update_responsibilities_threshold))
+ return -1;
reynolds_adaptation = false;
- else{
- PyErr_Format(PyExc_RuntimeError, "%s. The second argument must be a keyword argument.", Py_TYPE(self)->tp_name);
+ } else {
+ PyErr_Format(PyExc_RuntimeError, "%s. One of the two keyword arguments '%s' or '%s' must be present.", Py_TYPE(self)->tp_name, kwlist1[1], kwlist2[1]);
MAP_GMMTrainer_doc.print_usage();
return -1;
}
@@ -107,11 +111,11 @@ static int PyBobLearnEMMAPGMMTrainer_init_base_trainer(PyBobLearnEMMAPGMMTrainer
relevance_factor = aux;
else
alpha = aux;
-
-
- self->cxx.reset(new bob::learn::em::MAP_GMMTrainer(f(update_means), f(update_variances), f(update_weights),
- mean_var_update_responsibilities_threshold,
- reynolds_adaptation,relevance_factor, alpha, gmm_machine->cxx));
+
+
+ self->cxx.reset(new bob::learn::em::MAP_GMMTrainer(f(update_means), f(update_variances), f(update_weights),
+ mean_var_update_responsibilities_threshold,
+ reynolds_adaptation, relevance_factor, alpha, gmm_machine->cxx));
return 0;
}
@@ -202,12 +206,12 @@ PyObject* PyBobLearnEMMAPGMMTrainer_getRelevanceFactor(PyBobLearnEMMAPGMMTrainer
}
int PyBobLearnEMMAPGMMTrainer_setRelevanceFactor(PyBobLearnEMMAPGMMTrainerObject* self, PyObject* value, void*){
BOB_TRY
-
+
if(!PyBob_NumberCheck(value)){
PyErr_Format(PyExc_RuntimeError, "%s %s expects a double", Py_TYPE(self)->tp_name, relevance_factor.name());
return -1;
}
-
+
self->cxx->setRelevanceFactor(PyFloat_AS_DOUBLE(value));
return 0;
BOB_CATCH_MEMBER("relevance_factor could not be set", 0)
@@ -228,12 +232,12 @@ PyObject* PyBobLearnEMMAPGMMTrainer_getAlpha(PyBobLearnEMMAPGMMTrainerObject* se
}
int PyBobLearnEMMAPGMMTrainer_setAlpha(PyBobLearnEMMAPGMMTrainerObject* self, PyObject* value, void*){
BOB_TRY
-
+
if(!PyBob_NumberCheck(value)){
PyErr_Format(PyExc_RuntimeError, "%s %s expects a double", Py_TYPE(self)->tp_name, alpha.name());
return -1;
}
-
+
self->cxx->setAlpha(PyFloat_AS_DOUBLE(value));
return 0;
BOB_CATCH_MEMBER("alpha could not be set", 0)
@@ -241,7 +245,7 @@ int PyBobLearnEMMAPGMMTrainer_setAlpha(PyBobLearnEMMAPGMMTrainerObject* self, Py
-static PyGetSetDef PyBobLearnEMMAPGMMTrainer_getseters[] = {
+static PyGetSetDef PyBobLearnEMMAPGMMTrainer_getseters[] = {
{
alpha.name(),
(getter)PyBobLearnEMMAPGMMTrainer_getAlpha,
@@ -282,14 +286,14 @@ static PyObject* PyBobLearnEMMAPGMMTrainer_initialize(PyBobLearnEMMAPGMMTrainerO
char** kwlist = initialize.kwlist(0);
PyBobLearnEMGMMMachineObject* gmm_machine = 0;
- PyBlitzArrayObject* data = 0;
+ PyBlitzArrayObject* data = 0;
if (!PyArg_ParseTupleAndKeywords(args, kwargs, "O!|O&", kwlist, &PyBobLearnEMGMMMachine_Type, &gmm_machine,
&PyBlitzArray_Converter, &data)) return 0;
if(data != NULL)
auto data_ = make_safe(data);
self->cxx->initialize(*gmm_machine->cxx);
-
+
BOB_CATCH_MEMBER("cannot perform the initialize method", 0)
Py_RETURN_NONE;
@@ -324,16 +328,16 @@ static PyObject* PyBobLearnEMMAPGMMTrainer_eStep(PyBobLearnEMMAPGMMTrainerObject
auto data_ = make_safe(data);
- // perform check on the input
+ // perform check on the input
if (data->type_num != NPY_FLOAT64){
PyErr_Format(PyExc_TypeError, "`%s' only supports 64-bit float arrays for input array `%s`", Py_TYPE(self)->tp_name, eStep.name());
return 0;
- }
+ }
if (data->ndim != 2){
PyErr_Format(PyExc_TypeError, "`%s' only processes 2D arrays of float64 for `%s`", Py_TYPE(self)->tp_name, eStep.name());
return 0;
- }
+ }
if (data->shape[1] != (Py_ssize_t)gmm_machine->cxx->getNInputs() ) {
PyErr_Format(PyExc_TypeError, "`%s' 2D `input` array should have the shape [N, %" PY_FORMAT_SIZE_T "d] not [N, %" PY_FORMAT_SIZE_T "d] for `%s`", Py_TYPE(self)->tp_name, gmm_machine->cxx->getNInputs(), data->shape[1], eStep.name());
@@ -353,8 +357,8 @@ static PyObject* PyBobLearnEMMAPGMMTrainer_eStep(PyBobLearnEMMAPGMMTrainerObject
static auto mStep = bob::extension::FunctionDoc(
"mStep",
- "Performs a maximum a posteriori (MAP) update of the GMM:"
- "* parameters using the accumulated statistics in :py:class:`bob.learn.em.GMMBaseTrainer.m_ss` and the"
+ "Performs a maximum a posteriori (MAP) update of the GMM:"
+ "* parameters using the accumulated statistics in :py:class:`bob.learn.em.GMMBaseTrainer.m_ss` and the"
"* parameters of the prior model",
"",
true
@@ -369,7 +373,7 @@ static PyObject* PyBobLearnEMMAPGMMTrainer_mStep(PyBobLearnEMMAPGMMTrainerObject
char** kwlist = mStep.kwlist(0);
PyBobLearnEMGMMMachineObject* gmm_machine;
- PyBlitzArrayObject* data = 0;
+ PyBlitzArrayObject* data = 0;
if (!PyArg_ParseTupleAndKeywords(args, kwargs, "O!|O&", kwlist, &PyBobLearnEMGMMMachine_Type, &gmm_machine,
&PyBlitzArray_Converter, &data)) return 0;
@@ -473,6 +477,5 @@ bool init_BobLearnEMMAPGMMTrainer(PyObject* module)
// add the type to the module
Py_INCREF(&PyBobLearnEMMAPGMMTrainer_Type);
- return PyModule_AddObject(module, "_MAP_GMMTrainer", (PyObject*)&PyBobLearnEMMAPGMMTrainer_Type) >= 0;
+ return PyModule_AddObject(module, "MAP_GMMTrainer", (PyObject*)&PyBobLearnEMMAPGMMTrainer_Type) >= 0;
}
-
diff --git a/bob/learn/em/__MAP_gmm_trainer__.py b/bob/learn/em/__MAP_gmm_trainer__.py
deleted file mode 100644
index 322d88f..0000000
--- a/bob/learn/em/__MAP_gmm_trainer__.py
+++ /dev/null
@@ -1,48 +0,0 @@
-#!/usr/bin/env python
-# vim: set fileencoding=utf-8 :
-# Tiago de Freitas Pereira <tiago.pereira@idiap.ch>
-# Mon Jan 23 18:31:10 2015
-#
-# Copyright (C) 2011-2015 Idiap Research Institute, Martigny, Switzerland
-
-from ._library import _MAP_GMMTrainer
-import numpy
-
-# define the class
-class MAP_GMMTrainer(_MAP_GMMTrainer):
-
- def __init__(self, prior_gmm, update_means=True, update_variances=False, update_weights=False, **kwargs):
- """
- :py:class:`bob.learn.em.MAP_GMMTrainer` constructor
-
- Keyword Parameters:
- update_means
-
- update_variances
-
- update_weights
-
- prior_gmm
- A :py:class:`bob.learn.em.GMMMachine` to be adapted
- convergence_threshold
- Convergence threshold
- max_iterations
- Number of maximum iterations
- converge_by_likelihood
- Tells whether we compute log_likelihood as a convergence criteria, or not
- alpha
- Set directly the alpha parameter (Eq (14) from [Reynolds2000]_), ignoring zeroth order statistics as a weighting factor.
- relevance_factor
- If set the :py:class:`bob.learn.em.MAP_GMMTrainer.reynolds_adaptation` parameters, will apply the Reynolds Adaptation procedure. See Eq (14) from [Reynolds2000]_
- """
-
- if kwargs.get('alpha')!=None:
- alpha = kwargs.get('alpha')
- _MAP_GMMTrainer.__init__(self, prior_gmm,alpha=alpha, update_means=update_means, update_variances=update_variances,update_weights=update_weights)
- else:
- relevance_factor = kwargs.get('relevance_factor')
- _MAP_GMMTrainer.__init__(self, prior_gmm, relevance_factor=relevance_factor, update_means=update_means, update_variances=update_variances,update_weights=update_weights)
-
-
-# copy the documentation from the base class
-__doc__ = _MAP_GMMTrainer.__doc__
diff --git a/bob/learn/em/__init__.py b/bob/learn/em/__init__.py
index 445a35d..bf6f3fa 100644
--- a/bob/learn/em/__init__.py
+++ b/bob/learn/em/__init__.py
@@ -7,11 +7,9 @@ import bob.learn.linear
import bob.extension
bob.extension.load_bob_library('bob.learn.em', __file__)
-#from ._old_library import *
from ._library import *
from . import version
from .version import module as __version__
-from .__MAP_gmm_trainer__ import *
from .train import *
def ztnorm_same_value(vect_a, vect_b):
diff --git a/bob/learn/em/test/test_em.py b/bob/learn/em/test/test_em.py
index bd7ffc4..0dc6281 100644
--- a/bob/learn/em/test/test_em.py
+++ b/bob/learn/em/test/test_em.py
@@ -15,7 +15,7 @@ from bob.io.base.test_utils import datafile
from bob.learn.em import KMeansMachine, GMMMachine, KMeansTrainer, \
ML_GMMTrainer, MAP_GMMTrainer
-
+
import bob.learn.em
#, MAP_GMMTrainer
@@ -49,23 +49,23 @@ def test_gmm_ML_1():
# Trains a GMMMachine with ML_GMMTrainer
- ar = bob.io.base.load(datafile("faithful.torch3_f64.hdf5", __name__, path="../data/"))
+ ar = bob.io.base.load(datafile("faithful.torch3_f64.hdf5", __name__, path="../data/"))
gmm = loadGMM()
-
+
ml_gmmtrainer = ML_GMMTrainer(True, True, True)
#ml_gmmtrainer.train(gmm, ar)
bob.learn.em.train(ml_gmmtrainer, gmm, ar, convergence_threshold=0.001)
#config = bob.io.base.HDF5File(datafile('gmm_ML.hdf5", __name__), 'w')
#gmm.save(config)
-
+
gmm_ref = GMMMachine(bob.io.base.HDF5File(datafile('gmm_ML.hdf5', __name__, path="../data/")))
gmm_ref_32bit_debug = GMMMachine(bob.io.base.HDF5File(datafile('gmm_ML_32bit_debug.hdf5', __name__, path="../data/")))
gmm_ref_32bit_release = GMMMachine(bob.io.base.HDF5File(datafile('gmm_ML_32bit_release.hdf5', __name__, path="../data/")))
assert (gmm == gmm_ref) or (gmm == gmm_ref_32bit_release) or (gmm == gmm_ref_32bit_release)
-
+
def test_gmm_ML_2():
# Trains a GMMMachine with ML_GMMTrainer; compares to an old reference
@@ -86,11 +86,11 @@ def test_gmm_ML_2():
max_iter_gmm = 25
accuracy = 0.00001
ml_gmmtrainer = ML_GMMTrainer(True, True, True, prior)
-
+
# Run ML
#ml_gmmtrainer.train(gmm, ar)
bob.learn.em.train(ml_gmmtrainer, gmm, ar, max_iterations = max_iter_gmm, convergence_threshold=accuracy)
-
+
# Test results
# Load torch3vision reference
meansML_ref = bob.io.base.load(datafile('meansAfterML.hdf5', __name__, path="../data/"))
@@ -114,8 +114,8 @@ def test_gmm_MAP_1():
gmm = GMMMachine(bob.io.base.HDF5File(datafile("gmm_ML.hdf5", __name__, path="../data/")))
gmmprior = GMMMachine(bob.io.base.HDF5File(datafile("gmm_ML.hdf5", __name__, path="../data/")))
- map_gmmtrainer = MAP_GMMTrainer(update_means=True, update_variances=False, update_weights=False, prior_gmm=gmmprior, relevance_factor=4.)
-
+ map_gmmtrainer = MAP_GMMTrainer(update_means=True, update_variances=False, update_weights=False, prior_gmm=gmmprior, relevance_factor=4.)
+
#map_gmmtrainer.train(gmm, ar)
bob.learn.em.train(map_gmmtrainer, gmm, ar)
@@ -183,7 +183,7 @@ def test_gmm_MAP_3():
max_iter_gmm = 1
accuracy = 0.00001
map_factor = 0.5
- map_gmmtrainer = MAP_GMMTrainer(prior_gmm, alpha=map_factor, update_means=True, update_variances=False, update_weights=False, convergence_threshold=prior)
+ map_gmmtrainer = MAP_GMMTrainer(prior_gmm, alpha=map_factor, update_means=True, update_variances=False, update_weights=False, mean_var_update_responsibilities_threshold=accuracy)
#map_gmmtrainer.max_iterations = max_iter_gmm
#map_gmmtrainer.convergence_threshold = accuracy
@@ -192,7 +192,7 @@ def test_gmm_MAP_3():
# Train
#map_gmmtrainer.train(gmm, ar)
- bob.learn.em.train(map_gmmtrainer, gmm, ar, max_iterations = max_iter_gmm, convergence_threshold=accuracy)
+ bob.learn.em.train(map_gmmtrainer, gmm, ar, max_iterations = max_iter_gmm, convergence_threshold=prior)
# Test results
# Load torch3vision reference
--
GitLab