Migrate modelfun from boilerdata

.tools/requirements/requirements.txt

@@ -1,4 +1,5 @@
 # Latest working requirements of your package, should be ahead of pyproject.toml.
+dill==0.3.7
 IPython[notebook]==8.14.0
 pandas[hdf5,performance]==2.0.2
 pandas-stubs~=2.0.2

pyproject.toml

@@ -17,6 +17,7 @@ license = { file = "LICENSE" }
 requires-python = ">=3.11"
 classifiers = ["License :: OSI Approved :: MIT License"]
 dependencies = [
+    "dill>=0.3.7",
     "IPython[notebook]>=8.14.0",
     "pandas[hdf5,performance]>=2.0.2",
     "ploomber-engine>=0.0.30",

src/boilercore/modelfun.py

@@ -0,0 +1,38 @@
+import warnings
+from collections.abc import Callable
+from functools import wraps
+from pathlib import Path
+from typing import Any
+
+import dill
+import numpy as np
+
+
+def get_model(model: Path):
+    """Unpickle the model function for fitting data."""
+    file_bytes = Path(model).read_bytes()
+    with warnings.catch_warnings():
+        warnings.simplefilter("ignore", dill.UnpicklingWarning)
+        unpickled_model = dill.loads(file_bytes)
+    return unpickled_model.basic, fix_model(unpickled_model.for_ufloat)
+
+
+def fix_model(f) -> Callable[..., Any]:
+    """Fix edge-cases of lambdify where all inputs must be arrays.
+
+    See the notes section in the link below where it says, "However, in some cases
+    the generated function relies on the input being a numpy array."
+
+    https://docs.sympy.org/latest/modules/utilities/lambdify.html#sympy.utilities.lambdify.lambdify
+    """
+
+    @wraps(f)
+    def wrapper(*args, **kwargs):
+        result = f(
+            *(np.array(arg) for arg in args),
+            **{k: np.array(v) for k, v in kwargs.items()},
+        )
+
+        return result if result.size > 1 else result.item()
+
+    return wrapper