rmontanana/Pyclassifiers
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Fix fit & predict with discretized datasets

Browse Source
This commit is contained in:
Ricardo Montañana Gómez
2024-05-27 12:00:09 +02:00
parent 99c57d95f3
commit 235c345e87
5 changed files with 38 additions and 12 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
lib/catch2

Submodule lib/catch2 updated: cd8f97e6c7...4e8d92bf02

2
lib/json

Submodule lib/json updated: 377c767aa1...8c391e04fe

2
lib/mdlp

Submodule lib/mdlp updated: 5708dc3de9...236d1b2f8b

22
pyclfs/PyClassifier.cc
View File

@@ -21,10 +21,22 @@ namespace pywrap {
Xn = Xn.transpose();
return Xn;
}
np::ndarray tensorInt2numpy(torch::Tensor& X)
{
int m = X.size(0);
int n = X.size(1);
auto Xn = np::from_data(X.data_ptr(), np::dtype::get_builtin<int>(), bp::make_tuple(m, n), bp::make_tuple(sizeof(X.dtype()) * 2 * n, sizeof(X.dtype()) * 2), bp::object());
Xn = Xn.transpose();
//std::cout << "Transposed array:\n" << boost::python::extract<char const*>(boost::python::str(Xn)) << std::endl;
return Xn;
}
std::pair<np::ndarray, np::ndarray> tensors2numpy(torch::Tensor& X, torch::Tensor& y)
{
int n = X.size(1);
auto yn = np::from_data(y.data_ptr(), np::dtype::get_builtin<int32_t>(), bp::make_tuple(n), bp::make_tuple(sizeof(y.dtype()) * 2), bp::object());
if (X.dtype() == torch::kInt32) {
return { tensorInt2numpy(X), yn };
}
return { tensor2numpy(X), yn };
}
std::string PyClassifier::version()
@@ -65,8 +77,14 @@ namespace pywrap {
torch::Tensor PyClassifier::predict(torch::Tensor& X)
{
int dimension = X.size(1);
auto Xn = tensor2numpy(X);
CPyObject Xp = bp::incref(bp::object(Xn).ptr());
CPyObject Xp;
if (X.dtype() == torch::kInt32) {
auto Xn = tensorInt2numpy(X);
Xp = bp::incref(bp::object(Xn).ptr());
} else {
auto Xn = tensor2numpy(X);
Xp = bp::incref(bp::object(Xn).ptr());
}
PyObject* incoming = pyWrap->predict(id, Xp);
bp::handle<> handle(incoming);
bp::object object(handle);

22
tests/TestPythonClassifiers.cc
View File

@@ -13,7 +13,7 @@
#include "pyclfs/ODTE.h"
#include "TestUtils.h"
const std::string ACTUAL_VERSION = "1.0.4";
const std::string ACTUAL_VERSION = "1.0.5";
TEST_CASE("Test Python Classifiers score", "[PyClassifiers]")
{
@@ -60,19 +60,27 @@ TEST_CASE("Test Python Classifiers score", "[PyClassifiers]")
}
TEST_CASE("Classifiers features", "[PyClassifiers]")
{
auto raw = RawDatasets("iris", true);
auto raw = RawDatasets("iris", false);
auto clf = pywrap::STree();
clf.fit(raw.Xt, raw.yt, raw.featurest, raw.classNamet, raw.statest);
REQUIRE(clf.getNumberOfNodes() == 3);
REQUIRE(clf.getNumberOfEdges() == 2);
REQUIRE(clf.getNumberOfNodes() == 5);
REQUIRE(clf.getNumberOfEdges() == 3);
}
TEST_CASE("Get num features & num edges", "[PyClassifiers]")
{
auto raw = RawDatasets("iris", true);
auto raw = RawDatasets("iris", false);
auto clf = pywrap::ODTE();
clf.fit(raw.Xt, raw.yt, raw.featurest, raw.classNamet, raw.statest);
REQUIRE(clf.getNumberOfNodes() == 10);
REQUIRE(clf.getNumberOfEdges() == 10);
REQUIRE(clf.getNumberOfNodes() == 50);
REQUIRE(clf.getNumberOfEdges() == 30);
}
TEST_CASE("Classifier with discretized dataset", "[PyClassifiers]")
{
auto raw = RawDatasets("iris", true);
auto clf = pywrap::SVC();
clf.fit(raw.Xt, raw.yt, raw.featurest, raw.classNamet, raw.statest);
auto score = clf.score(raw.Xt, raw.yt);
REQUIRE(score == Catch::Approx(0.96667f).epsilon(raw.epsilon));
}
// TEST_CASE("XGBoost", "[PyClassifiers]")
// {