From 666eb5dbde039715691758308e2a2978917ba755 Mon Sep 17 00:00:00 2001
From: Tiago Freitas Pereira <tiagofrepereira@gmail.com>
Date: Thu, 31 Mar 2022 15:54:55 +0200
Subject: [PATCH] [precommit] Fixes
---
bob/learn/em/factor_analysis.py | 1 -
bob/learn/em/test/test_jfa_trainer.py | 13 ++-----------
doc/guide.rst | 10 +++++-----
doc/plot/plot_MAP.py | 3 ++-
4 files changed, 9 insertions(+), 18 deletions(-)
diff --git a/bob/learn/em/factor_analysis.py b/bob/learn/em/factor_analysis.py
index 275a8af..fc11ded 100644
--- a/bob/learn/em/factor_analysis.py
+++ b/bob/learn/em/factor_analysis.py
@@ -2,7 +2,6 @@
# @author: Tiago de Freitas Pereira
-from ast import Return
import logging
import numpy as np
diff --git a/bob/learn/em/test/test_jfa_trainer.py b/bob/learn/em/test/test_jfa_trainer.py
index 5d41c7c..f2b08d1 100644
--- a/bob/learn/em/test/test_jfa_trainer.py
+++ b/bob/learn/em/test/test_jfa_trainer.py
@@ -378,6 +378,8 @@ def test_ISVTrainAndEnrol():
gse2.sum_px = Fe[:, 1].reshape(2, 3)
gse = [gse1, gse2]
+ latent_z = it.enroll(gse, 5)
+ assert np.allclose(latent_z, z_ref, eps)
def test_ISVTrainAndEnrolWithNumpy():
@@ -406,17 +408,6 @@ def test_ISVTrainAndEnrolWithNumpy():
],
"float64",
)
- z_ref = np.array(
- [
- -0.079315777443826,
- 0.092702428248543,
- -0.342488761656616,
- -0.059922635809136,
- 0.133539981073604,
- 0.213118695516570,
- ],
- "float64",
- )
"""
Calls the train function
diff --git a/doc/guide.rst b/doc/guide.rst
index 51e308d..a94857f 100644
--- a/doc/guide.rst
+++ b/doc/guide.rst
@@ -203,7 +203,7 @@ Follow bellow an snippet on how to train a GMM using the MAP estimator.
... [ 0.5, -0.5, 0.2 ]])
>>> prior_gmm.weights = numpy.array([ 0.8, 0.5])
>>> # Creating the model for the adapted GMM, and setting the `prior_gmm` as the source GMM
- >>> # note that we have set `trainer="map"`, so we use the Maximum a posteriori estimator
+ >>> # note that we have set `trainer="map"`, so we use the Maximum a posteriori estimator
>>> adapted_gmm = bob.learn.em.GMMMachine(2, ubm=prior_gmm, trainer="map")
>>> # Training
>>> adapted_gmm = adapted_gmm.fit(data)
@@ -271,9 +271,9 @@ prior GMM.
... [1.2, 1.4, 1],
... [0.8, 1., 1]], dtype='float64')
>>> # Training a GMM with 2 Gaussians of dimension 3
- >>> prior_gmm = bob.learn.em.GMMMachine(2).fit(data)
+ >>> prior_gmm = bob.learn.em.GMMMachine(2).fit(data)
>>> # Creating the container
- >>> gmm_stats = prior_gmm.acc_statistics(data)
+ >>> gmm_stats = prior_gmm.acc_statistics(data)
>>> # Printing the responsibilities
>>> print(gmm_stats.n/gmm_stats.t)
[0.6 0.4]
@@ -352,7 +352,7 @@ The snippet bellow shows how to:
>>> model = isv_machine.enroll_with_array(enroll_data)
>>> print(model)
[[ 0.54 0.246 0.505 1.617 -0.791 0.746]]
-
+
>>> # Probing
>>> probe_data = np.array([[1.2, 0.1, 1.4], [0.5, 0.2, 0.3]])
>>> score = isv_machine.score_with_array(model, probe_data)
@@ -414,7 +414,7 @@ such session variability model.
>>> # Finally doing the JFA training with U and V subspaces with dimension of 2
>>> jfa_machine = bob.learn.em.JFAMachine(ubm, r_U=2, r_V=2).fit(gmm_stats, y)
- >>> print(jfa_machine.U)
+ >>> print(jfa_machine.U)
[[-0.069 -0.029]
[ 0.079 0.039]
[ 0.123 0.042]
diff --git a/doc/plot/plot_MAP.py b/doc/plot/plot_MAP.py
index f668c8c..24e12ca 100644
--- a/doc/plot/plot_MAP.py
+++ b/doc/plot/plot_MAP.py
@@ -1,8 +1,9 @@
import matplotlib.pyplot as plt
+import numpy as np
+
from sklearn.datasets import load_iris
import bob.learn.em
-import numpy as np
np.random.seed(10)
--
GitLab