feat: add VisdomXGBLogger with autolog for xgboost training

README.md

@@ -451,6 +451,34 @@ VisdomSklearnLogger.autolog(viz, env="sklearn_run")
 
 See `example/train_sklearn_example.py` for a full working example covering plain estimators and grid search.
 
+### XGBoost
+
+`visdom.loggers.VisdomXGBLogger` implements XGBoost's `TrainingCallback` protocol, plotting train/eval metrics to Visdom after every boosting round.
+
+There was no way to visualize XGBoost training runs in Visdom without manually attaching a `TrainingCallback` and wiring up `viz.line()` calls yourself. This adds opt-in auto-logging behind a single `autolog()` call, with no changes required to model, `train()`, or `fit()` code.
+
+```python
+import xgboost as xgb
+import visdom
+from visdom.loggers import VisdomXGBLogger
+
+viz = visdom.Visdom()
+VisdomXGBLogger.autolog(viz, env="xgb_run")
+
+booster = xgb.train(params, dtrain, evals=[(dtrain, "train"), (dval, "eval")])  # logged automatically
+clf = xgb.XGBClassifier().fit(X_train, y_train, eval_set=[(X_val, y_val)])      # logged automatically
+xgb.cv(params, dtrain, nfold=3)                                                 # logged automatically
+```
+
+Or attach a logger to a single run without patching anything:
+
+```python
+callback = VisdomXGBLogger(viz, env="xgb_run")
+booster = xgb.train(params, dtrain, evals=[(dtrain, "train"), (dval, "eval")], callbacks=[callback])
+```
+
+Each eval metric gets its own window with one trace per data name (`train`/`eval`), and `best_iteration`/`best_score` are logged as a text pane once training finishes. See `example/train_xgboost_example.py` for a full working example.
+
 ## Details
 <img src="https://user-images.githubusercontent.com/19650074/198747904-7a8a580f-851a-45fb-8f45-94e54a910ee2.png"/>
 

example/train_xgboost_example.py

@@ -0,0 +1,45 @@
+import xgboost as xgb
+from sklearn.datasets import make_classification
+from sklearn.model_selection import train_test_split
+
+import visdom
+from visdom.loggers import VisdomXGBLogger
+
+
+def main():
+    # synthetic classification: 500 samples, 20 features, 2 classes
+    X, y = make_classification(
+        n_samples=500,
+        n_features=20,
+        n_informative=10,
+        random_state=42,
+    )
+    X_train, X_test, y_train, y_test = train_test_split(
+        X, y, test_size=0.2, random_state=42
+    )
+
+    viz = visdom.Visdom()
+    VisdomXGBLogger.autolog(viz, env="xgboost_run")
+
+    # functional API -> train/eval logloss curves logged automatically
+    dtrain = xgb.DMatrix(X_train, label=y_train)
+    dtest = xgb.DMatrix(X_test, label=y_test)
+    booster = xgb.train(
+        {"objective": "binary:logistic", "eval_metric": "logloss"},
+        dtrain,
+        num_boost_round=50,
+        evals=[(dtrain, "train"), (dtest, "eval")],
+    )
+
+    # sklearn API -> same curves logged automatically via patched fit()
+    clf = xgb.XGBClassifier(n_estimators=50, eval_metric="logloss")
+    clf.fit(X_train, y_train, eval_set=[(X_test, y_test)])
+
+    preds = (booster.predict(dtest) > 0.5).astype(int)
+    accuracy = (preds == y_test).mean()
+    print("Booster accuracy: {:.4f}".format(accuracy))
+    print("XGBClassifier accuracy: {:.4f}".format(clf.score(X_test, y_test)))
+
+
+if __name__ == "__main__":
+    main()

py/visdom/loggers/__init__.py

@@ -9,3 +9,10 @@
 from visdom.loggers.sklearn import VisdomSklearnLogger
 
 __all__ = ["VisdomLogger", "VisdomSklearnLogger"]
+
+try:
+    from visdom.loggers.xgboost import VisdomXGBLogger
+
+    __all__.append("VisdomXGBLogger")
+except ImportError:
+    pass