add script to predict GP #90

scripts/predict_gp.py

@@ -0,0 +1,38 @@
+#!/usr/bin/env python
+
+import logging
+from pathlib import Path
+
+import numpy as np
+import pandas as pd
+import torch
+
+from swan.dataset import FingerprintsData, load_split_dataset
+from swan.modeller import GPModeller
+from swan.modeller.models import GaussianProcess
+from swan.utils.log_config import configure_logger
+
+# Starting logger
+configure_logger(Path("."))
+LOGGER = logging.getLogger(__name__)
+
+# Set float size default
+torch.set_default_dtype(torch.float32)
+
+partition = load_split_dataset()
+features, labels = [torch.from_numpy(getattr(partition, x).astype(np.float32)) for x in ("features_trainset", "labels_trainset")]
+model = GaussianProcess(features, labels.flatten())
+data = FingerprintsData(Path("tests/files/smiles.csv"), properties=None, sanitize=False)
+
+researcher = GPModeller(model, data, use_cuda=False, replace_state=False)
+# # If the labels are scaled you need to load the scaling functionality
+# researcher.data.load_scale()
+researcher.load_model("swan_chk.pt")
+
+fingers = data.fingerprints
+print("shape fingers: ", fingers.shape)
+predicted = researcher.predict(fingers)
+df = pd.DataFrame(
+    {"smiles": data.dataframe.smiles, "mean": predicted.mean, "lower": predicted.lower, "upper": predicted.upper})
+
+df.to_csv("predicted_values.csv")