240506 SP24 Course PR

docs/api/datasets/pyhealth.datasets.GPMDataset.rst

@@ -0,0 +1,17 @@
+pyhealth.datasets.GPMDataset
+===================================
+
+This class defines a PyTorch Dataset for processing proteomics data
+    for multiple species from the Global Proteomics Database at:
+    https://gpmdb.thegpm.org/thegpm-cgi/peptides_by_species.pl
+
+.. autoclass:: pyhealth.datasets.GPMDataset
+    :members:
+    :undoc-members:
+    :show-inheritance:
+
+
+
+
+
+

docs/requirements.txt

@@ -22,5 +22,5 @@ bokeh==3.0.1
 gspread==5.6.2
 google-cloud-storage==2.6.0
 oauth2client==4.1.3
-Jinja2==3.1.4
+Jinja2==3.1.3
 flask==2.2.5

examples/hf_mimic3_rnn.py

@@ -0,0 +1,46 @@
+from pyhealth.datasets import MIMIC3Dataset
+from pyhealth.datasets import split_by_patient, get_dataloader
+from pyhealth.models import RNN
+from pyhealth.tasks import hf_prediction_mimic3_fn
+from pyhealth.trainer import Trainer
+
+# STEP 1: load data
+base_dataset = MIMIC3Dataset(
+    root="/srv/local/data/physionet.org/files/mimiciii/1.4",
+    tables=["DIAGNOSES_ICD", "PROCEDURES_ICD", "PRESCRIPTIONS"],
+    code_mapping={"ICD9CM": "CCSCM"},
+    dev=True,
+    refresh_cache=False,
+)
+base_dataset.stat()
+
+# STEP 2: set task
+sample_dataset = base_dataset.set_task(hf_prediction_mimic3_fn)
+sample_dataset.stat()
+
+train_dataset, val_dataset, test_dataset = split_by_patient(
+    sample_dataset, [0.8, 0.1, 0.1]
+)
+train_dataloader = get_dataloader(train_dataset, batch_size=32, shuffle=True)
+val_dataloader = get_dataloader(val_dataset, batch_size=32, shuffle=False)
+test_dataloader = get_dataloader(test_dataset, batch_size=32, shuffle=False)
+
+# STEP 3: define model
+model = RNN(
+    dataset=sample_dataset,
+    feature_keys=["conditions", "procedures", "drugs"],
+    label_key="label",
+    mode="binary",
+)
+
+# STEP 4: define trainer
+trainer = Trainer(model=model)
+trainer.train(
+    train_dataloader=train_dataloader,
+    val_dataloader=val_dataloader,
+    epochs=50,
+    monitor="roc_auc",
+)
+
+# STEP 5: evaluate
+trainer.evaluate(test_dataloader)