splits

+#! /usr/bin/env python3
+
+"""
+    code to create train data split
+"""
+
+import pandas
+
+if __name__ == "__main__":
+    df = pandas.read_csv("data/train.csv")
+    for n in [1000, 2500, 5000]:
+        df_ = df.sample(n, random_state=0)
+        df_.to_csv(f"data/train_{n}.csv", index=False)