chore(): Update the XGBoost server

servers/xgboostserver/.gitignore

@@ -0,0 +1,2 @@
+test/model.bst
+test/model.json

servers/xgboostserver/README.md

@@ -0,0 +1,48 @@
+# How to run the Jupyter notebook
+
+## Create a virtual environment and use it in Jupyter Lab
+
+1. Create a virtual environment
+```bash
+python -m venv xgboost-env
+```
+
+2. Activate it
+```bash
+source xgboost-env/bin/activate
+```
+
+3. Install the Jupyter kernel using the Python interpreter in this environment
+```bash
+python -m ipykernel install --user \
+  --name xgboost-env \
+  --display-name "Python (xgboost-env)"
+```
+
+4. Install development dependencies
+```bash
+pip install -r requirements-dev.txt
+```
+
+5. Additionally, install the `seldon_core` package
+  * You can either install it with pip from pypi with `pip install seldon-core`
+  * Or you can install it from source by:
+    *  Navigating to the <root>/python folder and `make install`
+
+6. Install the package dependencies
+```bash
+pip install -r ./xgboostserver/requirements.txt
+```
+
+## Open Jupyter Lab and run the cells
+
+1. Open Jupyter Lab
+```bash
+jupyter lab
+```
+
+2. Navigate to the browser, then to the `test` folder and select the notebook
+
+3. From the top right, choose the Kernel called `Python (xgboost-env)`
+
+4. Run the cells one after the other

servers/xgboostserver/requirements-dev.txt

@@ -0,0 +1,4 @@
+pytest >= 6.2.4
+jupyterlab == 4.3.4
+ipython_genutils == 0.2.0
+ipykernel == 6.29.5

servers/xgboostserver/test/test_metadata_and_input.py

@@ -0,0 +1,44 @@
+import os
+import tempfile
+import numpy as np
+import pytest
+import yaml
+import xgboost as xgb
+from xgboostserver.XGBoostServer import XGBoostServer
+
+
+@pytest.fixture
+def model_uri():
+    with tempfile.TemporaryDirectory() as temp_dir:
+        X = np.random.rand(100, 5)
+        y = np.random.randint(2, size=100)
+        dtrain = xgb.DMatrix(X, label=y)
+        params = {'objective': 'binary:logistic', 'eval_metric': 'error'}
+        booster = xgb.train(params, dtrain, num_boost_round=10)
+        model_path = os.path.join(temp_dir, "model.json")
+        booster.save_model(model_path)
+        yield temp_dir
+
+
+def test_init_metadata(model_uri):
+    metadata = {"key": "value"}
+    metadata_path = os.path.join(model_uri, "metadata.yaml")
+    with open(metadata_path, "w") as f:
+        yaml.dump(metadata, f)
+
+    server = XGBoostServer(model_uri)
+
+    loaded_metadata = server.init_metadata()
+
+    # Assert that the loaded metadata matches the original metadata
+    assert loaded_metadata == metadata
+
+
+def test_predict_invalid_input(model_uri):
+    # Create an instance of XGBoostServer with the model URI
+    server = XGBoostServer(model_uri)
+    server.load()
+
+    X_test = np.random.rand(10, 3)  # Incorrect number of features
+    with pytest.raises(ValueError):
+        server.predict(X_test, names=[])

servers/xgboostserver/test/test_metadata_and_predict.py

@@ -0,0 +1,43 @@
+import os
+import tempfile
+import numpy as np
+import pytest
+import yaml
+import xgboost as xgb
+from xgboostserver.XGBoostServer import XGBoostServer
+
+
+@pytest.fixture
+def model_uri():
+    with tempfile.TemporaryDirectory() as temp_dir:
+        X = np.random.rand(100, 5)
+        y = np.random.randint(2, size=100)
+        dtrain = xgb.DMatrix(X, label=y)
+        params = {'objective': 'binary:logistic', 'eval_metric': 'error'}
+        booster = xgb.train(params, dtrain, num_boost_round=10)
+        model_path = os.path.join(temp_dir, "model.json")
+        booster.save_model(model_path)
+        yield temp_dir
+
+
+def test_init_metadata(model_uri):
+    metadata = {"key": "value"}
+    metadata_path = os.path.join(model_uri, "metadata.yaml")
+    with open(metadata_path, "w") as f:
+        yaml.dump(metadata, f)
+
+    server = XGBoostServer(model_uri)
+
+    loaded_metadata = server.init_metadata()
+
+    # Assert that the loaded metadata matches the original metadata
+    assert loaded_metadata == metadata
+
+
+def test_predict_invalid_input(model_uri):
+    # Create an instance of XGBoostServer with the model URI
+    server = XGBoostServer(model_uri)
+    server.load()
+    X_test = np.random.rand(10, 3)  # Incorrect number of features
+    with pytest.raises(ValueError):
+        server.predict(X_test, names=[])