| Line |
Branch |
Exec |
Source |
| 1 |
|
|
// |
| 2 |
|
|
// Copyright (c) 2016 CNRS |
| 3 |
|
|
// |
| 4 |
|
|
|
| 5 |
|
|
#include "pinocchio/multibody/joint/joint-composite.hpp" |
| 6 |
|
|
#include "pinocchio/algorithm/joint-configuration.hpp" |
| 7 |
|
|
#include "pinocchio/algorithm/kinematics.hpp" |
| 8 |
|
|
|
| 9 |
|
|
#include <iostream> |
| 10 |
|
|
#include <boost/test/unit_test.hpp> |
| 11 |
|
|
#include <boost/utility/binary.hpp> |
| 12 |
|
|
|
| 13 |
|
|
using namespace pinocchio; |
| 14 |
|
|
using namespace Eigen; |
| 15 |
|
|
|
| 16 |
|
|
template<typename JointModel> |
| 17 |
|
|
void test_joint_methods( |
| 18 |
|
|
const JointModelBase<JointModel> & jmodel, JointModelComposite & jmodel_composite); |
| 19 |
|
|
|
| 20 |
|
|
template<typename JointModel> |
| 21 |
|
✗ |
void test_joint_methods(const JointModelBase<JointModel> & jmodel) |
| 22 |
|
|
{ |
| 23 |
|
✗ |
JointModelComposite jmodel_composite(jmodel); |
| 24 |
|
✗ |
test_joint_methods(jmodel, jmodel_composite); |
| 25 |
|
|
} |
| 26 |
|
|
|
| 27 |
|
|
template<typename JointModel> |
| 28 |
|
✗ |
void test_joint_methods( |
| 29 |
|
|
const JointModelBase<JointModel> & jmodel, JointModelComposite & jmodel_composite) |
| 30 |
|
|
{ |
| 31 |
|
|
typedef typename JointModelBase<JointModel>::JointDataDerived JointData; |
| 32 |
|
|
|
| 33 |
|
✗ |
JointData jdata = jmodel.createData(); |
| 34 |
|
✗ |
JointDataComposite jdata_composite = jmodel_composite.createData(); |
| 35 |
|
|
|
| 36 |
|
✗ |
jmodel_composite.setIndexes(jmodel.id(), jmodel.idx_q(), jmodel.idx_v()); |
| 37 |
|
|
|
| 38 |
|
|
typedef typename JointModel::ConfigVector_t ConfigVector_t; |
| 39 |
|
|
typedef typename JointModel::TangentVector_t TangentVector_t; |
| 40 |
|
|
typedef typename LieGroup<JointModel>::type LieGroupType; |
| 41 |
|
|
|
| 42 |
|
|
typedef Eigen::Matrix<typename ConfigVector_t::Scalar, Eigen::Dynamic, 1, ConfigVector_t::Options> |
| 43 |
|
|
DynConfigVectorType; |
| 44 |
|
|
typedef DynConfigVectorType DynTangentVectorType; |
| 45 |
|
|
|
| 46 |
|
✗ |
ConfigVector_t ql(ConfigVector_t::Constant(jmodel.nq(), -M_PI)); |
| 47 |
|
✗ |
ConfigVector_t qu(ConfigVector_t::Constant(jmodel.nq(), M_PI)); |
| 48 |
|
✗ |
DynConfigVectorType q = LieGroupType().randomConfiguration(ql, qu); |
| 49 |
|
|
|
| 50 |
|
✗ |
BOOST_CHECK(jmodel.nv() == jmodel_composite.nv()); |
| 51 |
|
✗ |
BOOST_CHECK(jmodel.nq() == jmodel_composite.nq()); |
| 52 |
|
|
|
| 53 |
|
✗ |
jmodel.calc(jdata, q); |
| 54 |
|
✗ |
jmodel_composite.calc(jdata_composite, q); |
| 55 |
|
|
|
| 56 |
|
✗ |
BOOST_CHECK(jdata_composite.M.isApprox((SE3)jdata.M)); |
| 57 |
|
✗ |
BOOST_CHECK(jdata_composite.S.matrix().isApprox(jdata.S.matrix())); |
| 58 |
|
|
|
| 59 |
|
✗ |
q = LieGroupType().randomConfiguration(ql, qu); |
| 60 |
|
✗ |
DynTangentVectorType v = TangentVector_t::Random(jmodel.nv()); |
| 61 |
|
✗ |
jmodel.calc(jdata, q, v); |
| 62 |
|
✗ |
jmodel_composite.calc(jdata_composite, q, v); |
| 63 |
|
|
|
| 64 |
|
✗ |
BOOST_CHECK(jdata_composite.M.isApprox((SE3)jdata.M)); |
| 65 |
|
✗ |
BOOST_CHECK(jdata_composite.S.matrix().isApprox(jdata.S.matrix())); |
| 66 |
|
✗ |
BOOST_CHECK(jdata_composite.v.isApprox((Motion)jdata.v)); |
| 67 |
|
✗ |
BOOST_CHECK(jdata_composite.c.isApprox((Motion)jdata.c)); |
| 68 |
|
|
|
| 69 |
|
|
// TODO: Not yet checked |
| 70 |
|
|
// { |
| 71 |
|
|
// VectorXd q1(jmodel.random()); |
| 72 |
|
|
// jmodel.normalize(q1); |
| 73 |
|
|
// VectorXd q2(jmodel.random()); |
| 74 |
|
|
// jmodel.normalize(q2); |
| 75 |
|
|
// VectorXd v(VectorXd::Random(jmodel.nv())); |
| 76 |
|
|
// |
| 77 |
|
|
// BOOST_CHECK(jmodel_composite.integrate(q1,v).isApprox(jmodel.integrate(q1,v))); |
| 78 |
|
|
// |
| 79 |
|
|
// TangentVector_t v1 = jmodel_composite.difference(q1,q2); |
| 80 |
|
|
// TangentVector_t v2 = jmodel.difference(q1,q2); |
| 81 |
|
|
// |
| 82 |
|
|
// BOOST_CHECK(v1.isApprox(v2)); |
| 83 |
|
|
// |
| 84 |
|
|
// const double alpha = 0.2; |
| 85 |
|
|
// BOOST_CHECK(jmodel_composite.interpolate(q1,q2,alpha).isApprox(jmodel.interpolate(q1,q2,alpha))); |
| 86 |
|
|
// BOOST_CHECK(math::fabs(jmodel_composite.distance(q1,q2)-jmodel.distance(q1,q2))<= |
| 87 |
|
|
// NumTraits<double>::dummy_precision()); |
| 88 |
|
|
// } |
| 89 |
|
|
|
| 90 |
|
✗ |
Inertia::Matrix6 I0(Inertia::Random().matrix()); |
| 91 |
|
|
|
| 92 |
|
✗ |
Inertia::Matrix6 I1 = I0; |
| 93 |
|
✗ |
Inertia::Matrix6 I2 = I0; |
| 94 |
|
|
|
| 95 |
|
✗ |
const Eigen::VectorXd armature = |
| 96 |
|
✗ |
Eigen::VectorXd::Random(jmodel.nv()) + Eigen::VectorXd::Ones(jmodel.nv()); |
| 97 |
|
✗ |
jmodel.calc_aba(jdata, armature, I1, true); |
| 98 |
|
✗ |
jmodel_composite.calc_aba(jdata_composite, armature, I2, true); |
| 99 |
|
|
|
| 100 |
|
✗ |
double prec = 1e-10; // higher tolerance to errors due to possible numerical imprecisions |
| 101 |
|
|
|
| 102 |
|
✗ |
BOOST_CHECK(jdata.U.isApprox(jdata_composite.U, prec)); |
| 103 |
|
✗ |
BOOST_CHECK(jdata.Dinv.isApprox(jdata_composite.Dinv, prec)); |
| 104 |
|
✗ |
BOOST_CHECK(jdata.UDinv.isApprox(jdata_composite.UDinv, prec)); |
| 105 |
|
|
|
| 106 |
|
|
// Checking the inertia was correctly updated |
| 107 |
|
|
// We use isApprox as usual, except for the freeflyer, |
| 108 |
|
|
// where the correct result is exacly zero and isApprox would fail. |
| 109 |
|
|
// Only for this single case, we use the infinity norm of the difference |
| 110 |
|
✗ |
if (jmodel.shortname() == "JointModelFreeFlyer") |
| 111 |
|
✗ |
BOOST_CHECK((I1 - I2).lpNorm<Eigen::Infinity>() < prec); |
| 112 |
|
|
else |
| 113 |
|
✗ |
BOOST_CHECK(I1.isApprox(I2, prec)); |
| 114 |
|
|
|
| 115 |
|
|
// Test operator= |
| 116 |
|
✗ |
JointModelComposite jmodel_composite2 = jmodel_composite; |
| 117 |
|
✗ |
JointDataComposite jdata_composite2 = jmodel_composite2.createData(); |
| 118 |
|
|
|
| 119 |
|
✗ |
jmodel_composite2.calc(jdata_composite2, q, v); |
| 120 |
|
|
|
| 121 |
|
✗ |
Inertia::Matrix6 I3 = I0; |
| 122 |
|
✗ |
jmodel_composite2.calc_aba(jdata_composite2, armature, I3, true); |
| 123 |
|
|
|
| 124 |
|
✗ |
if (jmodel.shortname() == "JointModelFreeFlyer") |
| 125 |
|
✗ |
BOOST_CHECK((I3 - I2).lpNorm<Eigen::Infinity>() < prec); |
| 126 |
|
|
else |
| 127 |
|
✗ |
BOOST_CHECK(I3.isApprox(I2, prec)); |
| 128 |
|
|
|
| 129 |
|
✗ |
BOOST_CHECK(jmodel_composite2 == jmodel_composite); |
| 130 |
|
|
|
| 131 |
|
✗ |
BOOST_CHECK(jdata_composite2.U.isApprox(jdata_composite.U, prec)); |
| 132 |
|
✗ |
BOOST_CHECK(jdata_composite2.Dinv.isApprox(jdata_composite.Dinv, prec)); |
| 133 |
|
✗ |
BOOST_CHECK(jdata_composite2.UDinv.isApprox(jdata_composite.UDinv, prec)); |
| 134 |
|
|
|
| 135 |
|
|
// Test operator << |
| 136 |
|
✗ |
std::cout << "JointModelComposite operator<<:\n" << jmodel_composite << std::endl; |
| 137 |
|
|
|
| 138 |
|
|
// Test block operators |
| 139 |
|
✗ |
const int m = 100; |
| 140 |
|
✗ |
Data::Matrix6x mat(Data::Matrix6x::Ones(6, m)); |
| 141 |
|
✗ |
const Data::Matrix6x mat_const(Data::Matrix6x::Ones(6, m)); |
| 142 |
|
✗ |
Model::ConfigVectorType vec(Model::ConfigVectorType::Ones(m)); |
| 143 |
|
✗ |
const Model::ConfigVectorType vec_const(Model::ConfigVectorType::Ones(m)); |
| 144 |
|
|
|
| 145 |
|
✗ |
BOOST_CHECK(jmodel.jointConfigSelector(vec) == jmodel_composite.jointConfigSelector(vec)); |
| 146 |
|
✗ |
BOOST_CHECK( |
| 147 |
|
|
jmodel.jointConfigSelector(vec_const) == jmodel_composite.jointConfigSelector(vec_const)); |
| 148 |
|
|
|
| 149 |
|
✗ |
BOOST_CHECK(jmodel.jointVelocitySelector(vec) == jmodel_composite.jointVelocitySelector(vec)); |
| 150 |
|
✗ |
BOOST_CHECK( |
| 151 |
|
|
jmodel.jointVelocitySelector(vec_const) == jmodel_composite.jointVelocitySelector(vec_const)); |
| 152 |
|
|
|
| 153 |
|
✗ |
BOOST_CHECK(jmodel.jointCols(mat) == jmodel_composite.jointCols(mat)); |
| 154 |
|
✗ |
BOOST_CHECK(jmodel.jointCols(mat_const) == jmodel_composite.jointCols(mat_const)); |
| 155 |
|
|
} |
| 156 |
|
|
|
| 157 |
|
|
struct TestJointComposite |
| 158 |
|
|
{ |
| 159 |
|
|
|
| 160 |
|
|
template<typename JointModel> |
| 161 |
|
✗ |
void operator()(const JointModelBase<JointModel> &) const |
| 162 |
|
|
{ |
| 163 |
|
✗ |
JointModel jmodel; |
| 164 |
|
✗ |
jmodel.setIndexes(0, 0, 0); |
| 165 |
|
|
|
| 166 |
|
✗ |
test_joint_methods(jmodel); |
| 167 |
|
|
} |
| 168 |
|
|
|
| 169 |
|
|
// void operator()(const JointModelBase<JointModelComposite> &) const |
| 170 |
|
|
// { |
| 171 |
|
|
// JointModelComposite jmodel_composite; |
| 172 |
|
|
// jmodel_composite.addJoint(pinocchio::JointModelRX()); |
| 173 |
|
|
// jmodel_composite.addJoint(pinocchio::JointModelRY()); |
| 174 |
|
|
// jmodel_composite.setIndexes(0,0,0); |
| 175 |
|
|
// |
| 176 |
|
|
// test_joint_methods(jmodel_composite); |
| 177 |
|
|
// } |
| 178 |
|
|
|
| 179 |
|
✗ |
void operator()(const JointModelBase<JointModelRevoluteUnaligned> &) const |
| 180 |
|
|
{ |
| 181 |
|
✗ |
JointModelRevoluteUnaligned jmodel(1.5, 1., 0.); |
| 182 |
|
✗ |
jmodel.setIndexes(0, 0, 0); |
| 183 |
|
|
|
| 184 |
|
✗ |
test_joint_methods(jmodel); |
| 185 |
|
|
} |
| 186 |
|
|
|
| 187 |
|
✗ |
void operator()(const JointModelBase<JointModelPrismaticUnaligned> &) const |
| 188 |
|
|
{ |
| 189 |
|
✗ |
JointModelPrismaticUnaligned jmodel(1.5, 1., 0.); |
| 190 |
|
✗ |
jmodel.setIndexes(0, 0, 0); |
| 191 |
|
✗ |
test_joint_methods(jmodel); |
| 192 |
|
|
} |
| 193 |
|
|
}; |
| 194 |
|
|
|
| 195 |
|
|
BOOST_AUTO_TEST_SUITE(BOOST_TEST_MODULE) |
| 196 |
|
|
|
| 197 |
|
✗ |
BOOST_AUTO_TEST_CASE(test_basic) |
| 198 |
|
|
{ |
| 199 |
|
|
typedef boost::variant< |
| 200 |
|
|
JointModelRX, JointModelRY, JointModelRZ, JointModelRevoluteUnaligned, JointModelSpherical, |
| 201 |
|
|
JointModelSphericalZYX, JointModelPX, JointModelPY, JointModelPZ, JointModelPrismaticUnaligned, |
| 202 |
|
|
JointModelFreeFlyer, JointModelPlanar, JointModelTranslation, JointModelRUBX, JointModelRUBY, |
| 203 |
|
|
JointModelRUBZ> |
| 204 |
|
|
Variant; |
| 205 |
|
|
|
| 206 |
|
✗ |
boost::mpl::for_each<Variant::types>(TestJointComposite()); |
| 207 |
|
|
} |
| 208 |
|
|
|
| 209 |
|
✗ |
BOOST_AUTO_TEST_CASE(chain) |
| 210 |
|
|
{ |
| 211 |
|
✗ |
JointModelComposite jmodel_composite; |
| 212 |
|
✗ |
jmodel_composite.addJoint(JointModelRZ()) |
| 213 |
|
✗ |
.addJoint(JointModelRY(), SE3::Random()) |
| 214 |
|
✗ |
.addJoint(JointModelRX()); |
| 215 |
|
✗ |
BOOST_CHECK_MESSAGE(jmodel_composite.nq() == 3, "Chain did not work"); |
| 216 |
|
✗ |
BOOST_CHECK_MESSAGE(jmodel_composite.nv() == 3, "Chain did not work"); |
| 217 |
|
✗ |
BOOST_CHECK_MESSAGE(jmodel_composite.njoints == 3, "Chain did not work"); |
| 218 |
|
|
} |
| 219 |
|
|
|
| 220 |
|
✗ |
BOOST_AUTO_TEST_CASE(vsZYX) |
| 221 |
|
|
{ |
| 222 |
|
✗ |
JointModelSphericalZYX jmodel_spherical; |
| 223 |
|
✗ |
jmodel_spherical.setIndexes(0, 0, 0); |
| 224 |
|
|
|
| 225 |
|
✗ |
JointModelComposite jmodel_composite((JointModelRZ())); |
| 226 |
|
✗ |
jmodel_composite.addJoint(JointModelRY()); |
| 227 |
|
✗ |
jmodel_composite.addJoint(JointModelRX()); |
| 228 |
|
|
|
| 229 |
|
✗ |
test_joint_methods(jmodel_spherical, jmodel_composite); |
| 230 |
|
|
} |
| 231 |
|
|
|
| 232 |
|
✗ |
BOOST_AUTO_TEST_CASE(vsTranslation) |
| 233 |
|
|
{ |
| 234 |
|
✗ |
JointModelTranslation jmodel_translation; |
| 235 |
|
✗ |
jmodel_translation.setIndexes(0, 0, 0); |
| 236 |
|
|
|
| 237 |
|
✗ |
JointModelComposite jmodel_composite((JointModelPX())); |
| 238 |
|
✗ |
jmodel_composite.addJoint(JointModelPY()); |
| 239 |
|
✗ |
jmodel_composite.addJoint(JointModelPZ()); |
| 240 |
|
|
|
| 241 |
|
✗ |
test_joint_methods(jmodel_translation, jmodel_composite); |
| 242 |
|
|
} |
| 243 |
|
|
|
| 244 |
|
✗ |
BOOST_AUTO_TEST_CASE(test_recursive_variant) |
| 245 |
|
|
{ |
| 246 |
|
|
/// Create joint composite with two joints, |
| 247 |
|
✗ |
JointModelComposite jmodel_composite_two_rx((JointModelRX())); |
| 248 |
|
✗ |
jmodel_composite_two_rx.addJoint(JointModelRY()); |
| 249 |
|
|
|
| 250 |
|
|
/// Create Joint composite with three joints, and add a composite in it, to test the |
| 251 |
|
|
/// recursive_wrapper |
| 252 |
|
✗ |
JointModelComposite jmodel_composite_recursive((JointModelFreeFlyer())); |
| 253 |
|
✗ |
jmodel_composite_recursive.addJoint(JointModelPlanar()); |
| 254 |
|
✗ |
jmodel_composite_recursive.addJoint(jmodel_composite_two_rx); |
| 255 |
|
|
} |
| 256 |
|
|
|
| 257 |
|
✗ |
BOOST_AUTO_TEST_CASE(test_copy) |
| 258 |
|
|
{ |
| 259 |
|
✗ |
JointModelComposite jmodel_composite_planar((JointModelPX())); |
| 260 |
|
✗ |
jmodel_composite_planar.addJoint(JointModelPY()); |
| 261 |
|
✗ |
jmodel_composite_planar.addJoint(JointModelRZ()); |
| 262 |
|
✗ |
jmodel_composite_planar.setIndexes(0, 0, 0); |
| 263 |
|
|
|
| 264 |
|
✗ |
JointDataComposite jdata_composite_planar = jmodel_composite_planar.createData(); |
| 265 |
|
|
|
| 266 |
|
✗ |
VectorXd q1(VectorXd::Random(3)); |
| 267 |
|
✗ |
VectorXd q1_dot(VectorXd::Random(3)); |
| 268 |
|
|
|
| 269 |
|
✗ |
JointModelComposite model_copy = jmodel_composite_planar; |
| 270 |
|
✗ |
JointDataComposite data_copy = model_copy.createData(); |
| 271 |
|
|
|
| 272 |
|
✗ |
BOOST_CHECK_MESSAGE( |
| 273 |
|
|
model_copy.nq() == jmodel_composite_planar.nq(), "Test Copy Composite, nq are differents"); |
| 274 |
|
✗ |
BOOST_CHECK_MESSAGE( |
| 275 |
|
|
model_copy.nv() == jmodel_composite_planar.nv(), "Test Copy Composite, nv are differents"); |
| 276 |
|
|
|
| 277 |
|
✗ |
jmodel_composite_planar.calc(jdata_composite_planar, q1, q1_dot); |
| 278 |
|
✗ |
model_copy.calc(data_copy, q1, q1_dot); |
| 279 |
|
|
} |
| 280 |
|
|
|
| 281 |
|
✗ |
BOOST_AUTO_TEST_CASE(test_kinematics) |
| 282 |
|
|
{ |
| 283 |
|
✗ |
Model model; |
| 284 |
|
✗ |
JointModelComposite jmodel_composite; |
| 285 |
|
|
|
| 286 |
|
✗ |
SE3 config = SE3::Random(); |
| 287 |
|
✗ |
JointIndex parent = 0; |
| 288 |
|
|
|
| 289 |
|
✗ |
for (int i = 0; i < 10; i++) |
| 290 |
|
|
{ |
| 291 |
|
✗ |
parent = model.addJoint(parent, JointModelRX(), config, "joint"); |
| 292 |
|
✗ |
jmodel_composite.addJoint(JointModelRX(), config); |
| 293 |
|
|
|
| 294 |
|
✗ |
config.setRandom(); |
| 295 |
|
|
} |
| 296 |
|
|
|
| 297 |
|
✗ |
Data data(model); |
| 298 |
|
|
|
| 299 |
|
✗ |
Model model_c; |
| 300 |
|
✗ |
model_c.addJoint(0, jmodel_composite, SE3::Identity(), "joint"); |
| 301 |
|
|
|
| 302 |
|
✗ |
Data data_c(model_c); |
| 303 |
|
|
|
| 304 |
|
✗ |
BOOST_CHECK(model.nv == model_c.nv); |
| 305 |
|
✗ |
BOOST_CHECK(model.nq == model_c.nq); |
| 306 |
|
|
|
| 307 |
|
✗ |
VectorXd q(VectorXd::Random(model.nv)); |
| 308 |
|
✗ |
forwardKinematics(model, data, q); |
| 309 |
|
✗ |
forwardKinematics(model_c, data_c, q); |
| 310 |
|
|
|
| 311 |
|
✗ |
BOOST_CHECK(data.oMi.back().isApprox(data_c.oMi.back())); |
| 312 |
|
|
|
| 313 |
|
✗ |
q.setRandom(model.nq); |
| 314 |
|
✗ |
VectorXd v(VectorXd::Random(model.nv)); |
| 315 |
|
✗ |
forwardKinematics(model, data, q, v); |
| 316 |
|
✗ |
forwardKinematics(model_c, data_c, q, v); |
| 317 |
|
|
|
| 318 |
|
✗ |
BOOST_CHECK(data.oMi.back().isApprox(data_c.oMi.back())); |
| 319 |
|
✗ |
BOOST_CHECK(data.v.back().isApprox(data_c.v.back())); |
| 320 |
|
|
|
| 321 |
|
✗ |
q.setRandom(model.nq); |
| 322 |
|
✗ |
v.setRandom(model.nv); |
| 323 |
|
✗ |
VectorXd a(VectorXd::Random(model.nv)); |
| 324 |
|
✗ |
forwardKinematics(model, data, q, v, a); |
| 325 |
|
✗ |
forwardKinematics(model_c, data_c, q, v, a); |
| 326 |
|
|
|
| 327 |
|
✗ |
BOOST_CHECK(data.oMi.back().isApprox(data_c.oMi.back())); |
| 328 |
|
✗ |
BOOST_CHECK(data.v.back().isApprox(data_c.v.back())); |
| 329 |
|
✗ |
BOOST_CHECK(data.a.back().isApprox(data_c.a.back())); |
| 330 |
|
|
} |
| 331 |
|
|
|
| 332 |
|
|
BOOST_AUTO_TEST_SUITE_END() |
| 333 |
|
|
|