Split

clinops.split provides train/test splitting strategies that are correct for clinical ML. Standard sklearn.train_test_split is inappropriate for clinical data for two reasons:

1. Future leakage — random splits can put a patient's t+1 observation in training and t-1 in test.
2. Patient leakage — when the same patient appears in both train and test, the model memorises patient-specific patterns rather than generalising to new patients.

---

TemporalSplitter

Splits on a datetime cutoff. All rows before the cutoff go to train, all rows after go to test. The correct approach for any time-series clinical data.

from clinops.split import TemporalSplitter

# Explicit cutoff
result = TemporalSplitter(cutoff="2155-01-01", time_col="charttime").split(df)

# Auto-compute cutoff from the data
result = TemporalSplitter(train_frac=0.8, time_col="charttime").split(df)

print(result.summary())
# Train: 38,400 rows (80.0%)
# Test:   9,600 rows (20.0%)
# cutoff: 2155-01-01 00:00:00

train_df = result.train
test_df  = result.test

---

PatientSplitter

Ensures all rows for a given patient appear in the same split. Prevents data leakage in datasets where patients have multiple admissions or time-points.

from clinops.split import PatientSplitter

result = PatientSplitter(id_col="subject_id", test_size=0.2).split(df)

# Guaranteed: no patient appears in both splits
assert not set(result.train["subject_id"]) & set(result.test["subject_id"])

---

StratifiedPatientSplitter

Combines patient-level splitting with outcome stratification. Ensures the train/test outcome rate matches the population rate while respecting patient boundaries.

Critical for imbalanced clinical endpoints — in-hospital mortality is typically 5–15% in MIMIC.

from clinops.split import StratifiedPatientSplitter

result = StratifiedPatientSplitter(
    id_col="subject_id",
    outcome_col="hospital_expire_flag",
    test_size=0.2,
).split(df)

print(result.summary())
# Train: 32,000 rows (80.0%)
# Test:   8,000 rows (20.0%)
# population_outcome_rate: 0.0821
# train_outcome_rate:      0.0819
# test_outcome_rate:       0.0826

---

SplitResult

All splitters return a SplitResult with consistent fields:

Field	Type	Description
train	pd.DataFrame	Training split
test	pd.DataFrame	Test split
summary()	str	Human-readable split statistics

---

Choosing the right splitter

Scenario	Recommended splitter
Time-series data, no patient ID	TemporalSplitter
Multiple rows per patient, balanced outcome	PatientSplitter
Multiple rows per patient, rare outcome	StratifiedPatientSplitter