rmontanana/mdlp
1
1
Fork 0
mirror of https://github.com/rmontanana/mdlp.git synced 2025-08-17 00:15:59 +00:00
Code Issues Packages Projects Releases Wiki Activity

Fix conan (#10)

Browse Source 
* Fix debug conan build target

* Add viewcoverage and fix coverage generation

* Add more tests to cover new integrity checks

* Add tests to accomplish 100%

* Fix conan-create makefile target
This commit is contained in:
Ricardo Montañana Gómez
2025-07-02 20:09:34 +02:00
committed by GitHub
parent c1759ba1ce
commit 6d8b55a808
15 changed files with 322 additions and 165 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
src/BinDisc.cpp
View File

@@ -49,7 +49,7 @@ namespace mdlp {
// Note: y parameter is validated but not used in binning strategy
fit(X);
}
std::vector<precision_t> linspace(precision_t start, precision_t end, int num)
std::vector<precision_t> BinDisc::linspace(precision_t start, precision_t end, int num)
{
// Input validation
if (num < 2) {
@@ -77,7 +77,7 @@ namespace mdlp {
{
return std::max(lower, std::min(n, upper));
}
std::vector<precision_t> percentile(samples_t& data, const std::vector<precision_t>& percentiles)
std::vector<precision_t> BinDisc::percentile(samples_t& data, const std::vector<precision_t>& percentiles)
{
// Input validation
if (data.empty()) {

3
src/BinDisc.h
View File

@@ -23,6 +23,9 @@ namespace mdlp {
// y is included for compatibility with the Discretizer interface
void fit(samples_t& X_, labels_t& y) override;
void fit(samples_t& X);
protected:
std::vector<precision_t> linspace(precision_t start, precision_t end, int num);
std::vector<precision_t> percentile(samples_t& data, const std::vector<precision_t>& percentiles);
private:
void fit_uniform(const samples_t&);
void fit_quantile(const samples_t&);

10
src/CPPFImdlp.h
View File

@@ -39,8 +39,8 @@ namespace mdlp {
size_t getCandidate(size_t, size_t);
size_t compute_max_num_cut_points() const;
pair<precision_t, size_t> valueCutPoint(size_t, size_t, size_t);
private:
inline precision_t safe_X_access(size_t idx) const {
inline precision_t safe_X_access(size_t idx) const
{
if (idx >= indices.size()) {
throw std::out_of_range("Index out of bounds for indices array");
}
@@ -50,7 +50,8 @@ namespace mdlp {
}
return X[real_idx];
}
inline label_t safe_y_access(size_t idx) const {
inline label_t safe_y_access(size_t idx) const
{
if (idx >= indices.size()) {
throw std::out_of_range("Index out of bounds for indices array");
}
@@ -60,7 +61,8 @@ namespace mdlp {
}
return y[real_idx];
}
inline size_t safe_subtract(size_t a, size_t b) const {
inline size_t safe_subtract(size_t a, size_t b) const
{
if (b > a) {
throw std::underflow_error("Subtraction would cause underflow");
}

17
src/Discretizer.cpp
View File

@@ -17,7 +17,7 @@ namespace mdlp {
if (cutPoints.size() < 2) {
throw std::runtime_error("Discretizer not fitted yet or no valid cut points found");
}
discretizedData.clear();
discretizedData.reserve(data.size());
// CutPoints always have at least two items
@@ -40,9 +40,6 @@ namespace mdlp {
void Discretizer::fit_t(const torch::Tensor& X_, const torch::Tensor& y_)
{
// Validate tensor properties for security
if (!X_.is_contiguous() || !y_.is_contiguous()) {
throw std::invalid_argument("Tensors must be contiguous");
}
if (X_.sizes().size() != 1 || y_.sizes().size() != 1) {
throw std::invalid_argument("Only 1D tensors supported");
}
@@ -58,7 +55,7 @@ namespace mdlp {
if (X_.numel() == 0) {
throw std::invalid_argument("Tensors cannot be empty");
}
auto num_elements = X_.numel();
samples_t X(X_.data_ptr<precision_t>(), X_.data_ptr<precision_t>() + num_elements);
labels_t y(y_.data_ptr<int>(), y_.data_ptr<int>() + num_elements);
@@ -67,9 +64,6 @@ namespace mdlp {
torch::Tensor Discretizer::transform_t(const torch::Tensor& X_)
{
// Validate tensor properties for security
if (!X_.is_contiguous()) {
throw std::invalid_argument("Tensor must be contiguous");
}
if (X_.sizes().size() != 1) {
throw std::invalid_argument("Only 1D tensors supported");
}
@@ -79,7 +73,7 @@ namespace mdlp {
if (X_.numel() == 0) {
throw std::invalid_argument("Tensor cannot be empty");
}
auto num_elements = X_.numel();
samples_t X(X_.data_ptr<precision_t>(), X_.data_ptr<precision_t>() + num_elements);
auto result = transform(X);
@@ -88,9 +82,6 @@ namespace mdlp {
torch::Tensor Discretizer::fit_transform_t(const torch::Tensor& X_, const torch::Tensor& y_)
{
// Validate tensor properties for security
if (!X_.is_contiguous() || !y_.is_contiguous()) {
throw std::invalid_argument("Tensors must be contiguous");
}
if (X_.sizes().size() != 1 || y_.sizes().size() != 1) {
throw std::invalid_argument("Only 1D tensors supported");
}
@@ -106,7 +97,7 @@ namespace mdlp {
if (X_.numel() == 0) {
throw std::invalid_argument("Tensors cannot be empty");
}
auto num_elements = X_.numel();
samples_t X(X_.data_ptr<precision_t>(), X_.data_ptr<precision_t>() + num_elements);
labels_t y(y_.data_ptr<int>(), y_.data_ptr<int>() + num_elements);