baybe.simulation.lookup.look_up_targets

baybe.simulation.lookup.look_up_targets(queries: DataFrame, targets: Collection[Target], lookup: DataFrame | Callable | None, impute_mode: Literal['error', 'worst', 'best', 'mean', 'random', 'ignore'] = 'error')

Add/fill target values in a dataframe using a lookup mechanism.

Note

This does not create a new dataframe but modifies queries in-place.

Parameters:

• queries (DataFrame) – The dataframe to be modified. Its content must be compatible with the chosen lookup mechanism.

• targets (Collection[Target]) – The targets whose values are to be looked up.

• lookup (DataFrame | Callable | None) –

  The lookup mechanism. Can be one of the following choices:

  • A dataframe mapping rows of queries to the corresponding target values. That is, it must contain the same columns as queries plus one additional column for each of the given target.

  • A callable, providing target values for each row of queries.

  • None. Produces fake values for all targets.

• impute_mode (Literal['error', 'worst', 'best', 'mean', 'random', 'ignore']) –

  Specifies how a missing lookup will be handled. Only relevant for dataframe lookups. Can be one of the following choices:

  • "error": An error will be thrown.

  • "worst": Imputes the worst available value for each target.

  • "best": Imputes the best available value for each target.

  • "mean": Imputes the mean value for each target.

  • "random": A random row will be used for the lookup.

Raises:

ValueError – If an unsupported lookup mechanism is provided.

Return type:

None

Example

>>> import pandas as pd
>>> from baybe.targets.numerical import NumericalTarget
>>> from baybe.simulation.lookup import look_up_targets
>>>
>>> targets = [NumericalTarget("target", "MAX")]
>>> df = pd.DataFrame({"x": [1, 2, 3]})
>>> lookup_df = pd.DataFrame({"x": [1, 2], "target": [10, 20]})
>>> look_up_targets(df, targets, lookup_df, impute_mode="mean")
>>> print(df)
   x  target
0  1    10.0
1  2    20.0
2  3    15.0