Add CLI binaries for statistical models and serialization for CITrees by Copilot · Pull Request #45 · hdclark/Ygor

Copilot · 2026-03-25T06:31:06Z

Expose stochastic forests, conditional forests, and conditional inference trees to end-users via command-line binaries, without requiring C++ development.

New binaries

Six binaries, all prefixed ygor_ and installed to CMAKE_INSTALL_BINDIR:

Binary	Purpose
`ygor_stochastic_forest_train`	Train stochastic forest (Breiman 2001)
`ygor_stochastic_forest_predict`	Predict with trained stochastic forest
`ygor_conditional_forest_train`	Train conditional random forest (Strobl et al. 2007/2008)
`ygor_conditional_forest_predict`	Predict with trained conditional forest
`ygor_ci_tree_train`	Train conditional inference tree (Hothorn et al. 2006)
`ygor_ci_tree_predict`	Predict with trained CI tree

All binaries use ArgumentHandler for CLI parsing, accept CSV/TSV with auto-detection (tab presence), and support --header to skip a header row. All model hyperparameters are exposed as CLI flags.

# Train with gini importance, save model
ygor_stochastic_forest_train -i data.csv -o model.txt -H -I gini -t 200 -d 15

# Predict
ygor_stochastic_forest_predict -m model.txt -i features.csv -H

Training binaries: last column = response, preceding columns = features. Displays feature importances post-fit when applicable. Prediction binaries: all columns = features, outputs one prediction per row to stdout.

CITrees serialization

ConditionalInferenceTrees lacked write_to/read_from—added following the identical pattern used by StochasticForests and ConditionalRandomForests (versioned header, parameter block, recursive tree node serialization with L/I node types, RAII precision guard).

Build

Updated binaries/CMakeLists.txt with all six targets linked against ygor, m, Threads::Threads and added to the install target.

💡 You can make Copilot smarter by setting up custom instructions, customizing its development environment and configuring Model Context Protocol (MCP) servers. Learn more Copilot coding agent tips in the docs.

Add command-line utilities for stochastic forests, conditional random forests, and conditional inference trees (train + predict for each). All binaries use ArgumentHandler for CLI args, support CSV/TSV input with auto-detection, and follow existing repository conventions. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

- Add write_to/read_from serialization to ConditionalInferenceTrees - Create ygor_stochastic_forest_train and ygor_stochastic_forest_predict - Create ygor_conditional_forest_train and ygor_conditional_forest_predict - Create ygor_ci_tree_train and ygor_ci_tree_predict - Update binaries/CMakeLists.txt to build and install all new binaries Co-authored-by: hdclark <934858+hdclark@users.noreply.github.com> Agent-Logs-Url: https://github.com/hdclark/Ygor/sessions/a5269fa4-9646-45da-9023-4ba7802d958c

…ader flag Co-authored-by: hdclark <934858+hdclark@users.noreply.github.com> Agent-Logs-Url: https://github.com/hdclark/Ygor/sessions/a5269fa4-9646-45da-9023-4ba7802d958c

Copilot

Pull request overview

This PR exposes Ygor’s statistical tree/forest models to end-users via six new command-line binaries and adds text-stream serialization support to Stats::ConditionalInferenceTrees, aligning it with the existing serialization patterns used by the forest models.

Changes:

Added write_to() / read_from() (and recursive node serialization helpers) for ConditionalInferenceTrees.
Introduced six new ygor_* CLI binaries to train/predict with stochastic forests, conditional random forests, and CI trees from CSV/TSV input.
Updated binaries/CMakeLists.txt to build, link, and install the new executables.

Reviewed changes

Copilot reviewed 9 out of 9 changed files in this pull request and generated 11 comments.

Show a summary per file

File	Description
src/YgorStatsCITrees.h	Declares CI tree serialization APIs and helpers.
src/YgorStatsCITrees.cc	Implements CI tree serialization/deserialization logic.
binaries/Ygor_Stochastic_Forest_Train.cc	New CLI trainer for stochastic forests (CSV/TSV ingestion + model save).
binaries/Ygor_Stochastic_Forest_Predict.cc	New CLI predictor for stochastic forests (model load + per-row predictions).
binaries/Ygor_Conditional_Forest_Train.cc	New CLI trainer for conditional random forests (hyperparameters + model save).
binaries/Ygor_Conditional_Forest_Predict.cc	New CLI predictor for conditional random forests.
binaries/Ygor_CI_Tree_Train.cc	New CLI trainer for conditional inference trees (hyperparameters + model save).
binaries/Ygor_CI_Tree_Predict.cc	New CLI predictor for conditional inference trees.
binaries/CMakeLists.txt	Adds build/link/install rules for the six new CLI targets.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2026-03-25T06:59:36Z

src/YgorStatsCITrees.cc

+        }else if(node_type == "I"){
+            node->is_leaf = false;
+            is >> node->split_feature;
+            std::string thresh_str;
+            is >> thresh_str;
+            if(is.fail()) return nullptr;
+            node->split_threshold = static_cast<T>(std::stold(thresh_str));
+            node->left = read_tree_node(is);
+            node->right = read_tree_node(is);
+            if(!node->left || !node->right) return nullptr;


read_tree_node() accepts split_feature from the stream without validating it against n_features_trained. A malformed/hostile model file could deserialize a negative or out-of-range feature index, which can later trigger out-of-bounds access in predict_tree() (via x.read_coeff(0, node->split_feature)). Add a bounds check during deserialization (after n_features_trained is read) and fail the load if the index is invalid.

Copilot · 2026-03-25T06:59:37Z

src/YgorStatsCITrees.cc

+
+    is >> label >> this->random_seed;
+    if(is.fail() || label != "random_seed") return false;
+


read_from() assigns serialized hyperparameters directly to members without re-validating the invariants enforced by the constructor (e.g., max_depth > 0, min_samples_split >= 2, alpha in (0,1), n_permutations > 0). This can leave the object in an invalid state after loading an incorrect/corrupted model file. Consider validating these ranges immediately after parsing and returning false if any are invalid.

Suggested change

// Validate hyperparameter invariants after deserialization.

if(this->max_depth <= 0){

return false;

}

if(this->min_samples_split < 2){

return false;

}

if(this->alpha <= static_cast<T>(0) || static_cast<T>(1) <= this->alpha){

return false;

}

if(this->n_permutations <= 0){

return false;

}

Copilot · 2026-03-25T06:59:37Z

binaries/Ygor_Conditional_Forest_Train.cc

+        std::stringstream ss(line);
+        std::string token;
+        while(std::getline(ss, token, delimiter)){
+            vals.push_back(std::stod(token));
+        }


CSV/TSV numeric parsing uses std::stod(token) without handling std::invalid_argument/std::out_of_range. If the input contains a non-numeric token (including empty fields), this will throw and terminate the program without a clear, row/column-specific error. Wrap conversions in a try/catch and rethrow a runtime_error that includes row/column context (and consider trimming whitespace).

Copilot · 2026-03-25T06:59:37Z

binaries/Ygor_CI_Tree_Train.cc

+        std::stringstream ss(line);
+        std::string token;
+        while(std::getline(ss, token, delimiter)){
+            vals.push_back(std::stod(token));
+        }


CSV/TSV numeric parsing uses std::stod(token) without handling std::invalid_argument/std::out_of_range. If the input contains a non-numeric token (including empty fields), this will throw and terminate the program without a clear, row/column-specific error. Wrap conversions in a try/catch and rethrow a runtime_error that includes row/column context (and consider trimming whitespace).

Copilot · 2026-03-25T06:59:37Z

binaries/Ygor_CI_Tree_Predict.cc

+        std::stringstream ss(line);
+        std::string token;
+        while(std::getline(ss, token, delimiter)){
+            vals.push_back(std::stod(token));
+        }


CSV/TSV numeric parsing uses std::stod(token) without handling std::invalid_argument/std::out_of_range. If the input contains a non-numeric token (including empty fields), this will throw and terminate the program without a clear, row/column-specific error. Wrap conversions in a try/catch and rethrow a runtime_error that includes row/column context (and consider trimming whitespace).

Copilot · 2026-03-25T06:59:38Z

src/YgorStatsCITrees.cc

+bool Stats::ConditionalInferenceTrees<T>::write_to(std::ostream &os) const {
+    const auto original_precision = os.precision();
+    os.precision( std::numeric_limits<T>::max_digits10 );
+


write_to()/read_from() add a new persistence format for ConditionalInferenceTrees, but there are currently no unit tests exercising round-trip correctness or invalid-input rejection (unlike StochasticForests and ConditionalRandomForests, which have serialization tests under tests2/). Adding tests in tests2/YgorStatsCITrees.cc for round-tripping predictions and for rejecting malformed streams would help prevent format regressions.

Copilot · 2026-03-25T06:59:38Z

binaries/Ygor_Stochastic_Forest_Train.cc

+            min_samples_split = std::stoll(optarg);
+        }));
+    arger.push_back(std::make_tuple(2, 'f', "max-features", true, "<int>",
+        "Maximum features per split; -1 for all (default: -1).",


The --max-features help text says "-1 for all", but Stats::StochasticForests treats max_features <= 0 as sqrt(n_features) (see library docs/implementation). This makes the CLI misleading and will produce different behavior than users expect. Update the flag semantics and/or help text (e.g., document <=0 as sqrt, and provide an explicit way to request "all features").

Suggested change

"Maximum features per split; -1 for all (default: -1).",

"Maximum features per split; <=0 uses sqrt(number of features) (default: -1).",

Copilot · 2026-03-25T06:59:38Z

binaries/Ygor_Stochastic_Forest_Train.cc

+        std::stringstream ss(line);
+        std::string token;
+        while(std::getline(ss, token, delimiter)){
+            vals.push_back(std::stod(token));
+        }


CSV/TSV numeric parsing uses std::stod(token) without handling std::invalid_argument/std::out_of_range. If the input contains a non-numeric token (including empty fields), this will throw and terminate the program without a clear, row/column-specific error. Wrap conversions in a try/catch and rethrow a runtime_error that includes row/column context (and consider trimming whitespace).

Copilot · 2026-03-25T06:59:38Z

binaries/Ygor_Stochastic_Forest_Predict.cc

+        std::stringstream ss(line);
+        std::string token;
+        while(std::getline(ss, token, delimiter)){
+            vals.push_back(std::stod(token));
+        }


CSV/TSV numeric parsing uses std::stod(token) without handling std::invalid_argument/std::out_of_range. If the input contains a non-numeric token (including empty fields), this will throw and terminate the program without a clear, row/column-specific error. Wrap conversions in a try/catch and rethrow a runtime_error that includes row/column context (and consider trimming whitespace).

Copilot · 2026-03-25T06:59:38Z

binaries/Ygor_Conditional_Forest_Predict.cc

+        std::stringstream ss(line);
+        std::string token;
+        while(std::getline(ss, token, delimiter)){
+            vals.push_back(std::stod(token));
+        }


CSV/TSV numeric parsing uses std::stod(token) without handling std::invalid_argument/std::out_of_range. If the input contains a non-numeric token (including empty fields), this will throw and terminate the program without a clear, row/column-specific error. Wrap conversions in a try/catch and rethrow a runtime_error that includes row/column context (and consider trimming whitespace).

Initial plan

04de628

Copilot AI assigned Copilot and hdclark Mar 25, 2026

Copilot started work on behalf of hdclark March 25, 2026 06:31 View session

Copilot AI and others added 3 commits March 25, 2026 06:44

Refactor CSV/TSV header parsing to avoid confusing mutation of has_he…

c57bb91

…ader flag Co-authored-by: hdclark <934858+hdclark@users.noreply.github.com> Agent-Logs-Url: https://github.com/hdclark/Ygor/sessions/a5269fa4-9646-45da-9023-4ba7802d958c

Copilot AI changed the title ~~[WIP] Add command line binaries for statistical models~~ Add CLI binaries for statistical models and serialization for CITrees Mar 25, 2026

Copilot AI requested a review from hdclark March 25, 2026 06:52

Copilot finished work on behalf of hdclark March 25, 2026 06:52

hdclark marked this pull request as ready for review March 25, 2026 06:53

Copilot AI review requested due to automatic review settings March 25, 2026 06:53

Copilot started reviewing on behalf of hdclark March 25, 2026 06:54 View session

Copilot AI reviewed Mar 25, 2026

View reviewed changes

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add CLI binaries for statistical models and serialization for CITrees#45

Add CLI binaries for statistical models and serialization for CITrees#45
Copilot wants to merge 4 commits intomasterfrom
copilot/add-command-line-binaries-for-models

Copilot AI commented Mar 25, 2026 •

edited

Loading

Uh oh!

Copilot AI left a comment

Uh oh!

Copilot AI Mar 25, 2026

Uh oh!

Copilot AI Mar 25, 2026

Uh oh!

Copilot AI Mar 25, 2026

Uh oh!

Copilot AI Mar 25, 2026

Uh oh!

Copilot AI Mar 25, 2026

Uh oh!

Copilot AI Mar 25, 2026

Uh oh!

Copilot AI Mar 25, 2026

Uh oh!

Copilot AI Mar 25, 2026

Uh oh!

Copilot AI Mar 25, 2026

Uh oh!

Copilot AI Mar 25, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants


		is >> label >> this->random_seed;
		if(is.fail() \|\| label != "random_seed") return false;

+    // Validate hyperparameter invariants after deserialization.
+    if(this->max_depth <= 0){
+        return false;
+    }
+    if(this->min_samples_split < 2){
+        return false;
+    }
+    if(this->alpha <= static_cast<T>(0) || static_cast<T>(1) <= this->alpha){
+        return false;
+    }
+    if(this->n_permutations <= 0){
+        return false;
+    }

	"Maximum features per split; -1 for all (default: -1).",
	"Maximum features per split; <=0 uses sqrt(number of features) (default: -1).",

Conversation

Copilot AI commented Mar 25, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

New binaries

CITrees serialization

Build

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Copilot AI Mar 25, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 25, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 25, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 25, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 25, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 25, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 25, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 25, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 25, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 25, 2026

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Copilot AI commented Mar 25, 2026 •

edited

Loading