From 1668225dc57156dea15a7e21fb7e7db591f2419d Mon Sep 17 00:00:00 2001
From: John Kitchin <jkitchin@andrew.cmu.edu>
Date: Tue, 9 Jul 2024 10:30:36 -0400
Subject: [PATCH] add sklearn__init, lruq and updates on lhc

---
 pycse/sklearn/__init__.py |  0
 pycse/sklearn/lhc.py      |  4 +++-
 pycse/sklearn/lr_uq.py    | 22 ++++++++++++++++++++++
 3 files changed, 25 insertions(+), 1 deletion(-)
 create mode 100644 pycse/sklearn/__init__.py
 create mode 100644 pycse/sklearn/lr_uq.py

diff --git a/pycse/sklearn/__init__.py b/pycse/sklearn/__init__.py
new file mode 100644
index 00000000..e69de29b
diff --git a/pycse/sklearn/lhc.py b/pycse/sklearn/lhc.py
index ae114d6f..15124f41 100644
--- a/pycse/sklearn/lhc.py
+++ b/pycse/sklearn/lhc.py
@@ -147,7 +147,9 @@ def anova(self):
 
         table = np.vstack([ns.index.values, ns.values, ns > fc]).T
 
-        return pd.DataFrame(table, columns=[f"{self.y} effect", "F-score", "Significant"])
+        return pd.DataFrame(
+            table, columns=[f"{self.y} effect", f"F-score (fc={fc:1.1f})", "Significant"]
+        )
 
     def predict(self, args):
         """Predict the response for ARGS. ARGS is a list of labels in order
diff --git a/pycse/sklearn/lr_uq.py b/pycse/sklearn/lr_uq.py
new file mode 100644
index 00000000..86a105b6
--- /dev/null
+++ b/pycse/sklearn/lr_uq.py
@@ -0,0 +1,22 @@
+"""
+A Linear regressor with uncertainty quantification.
+"""
+import numpy as np
+from pycse import regress
+from pycse import predict as _predict
+from sklearn.base import BaseEstimator, RegressorMixin
+
+
+class LinearRegressionUQ(BaseEstimator, RegressorMixin):
+    def fit(self, X, y):
+        self.xtrain = np.array(X)
+        self.ytrain = np.array(y)
+        self.coefs_, self.pars_cint, self.pars_se = regress(X, y, rcond=None)
+        return self
+
+    def predict(self, X, return_std=False):
+        y, _, se = _predict(self.xtrain, self.ytrain, self.coefs_, X)
+        if return_std:
+            return y, se
+        else:
+            return y