adaptive.run_adaptive_search

run_adaptive_search(
    command,
    input_cols,
    outcome_col,
    ranges,
    existing_data=None,
    n_start=20,
    n_step=10,
    max_iter=5,
    max_hours=None,
    input_file='sim_input.csv',
    output_file='sim_output.csv',
)

Orchestrates the Active Learning loop on raw DataFrames.

Parameters

Name	Type	Description	Default
command	str	Solver command template.	required
input_cols	List[str]	Input names.	required
outcome_col	str	Outcome name.	required
ranges	Dict	Input bounds.	required
existing_data	pd.DataFrame	Start data.	None
n_start	int	Init batch size.	20
n_step	int	Points added per refinement step.	10
max_iter	int	Max loops.	5
max_hours	float	Physical time limit in hours.	None
input_file	str	Temporary file for solver input.	'sim_input.csv'
output_file	str	Temporary file for solver output.	'sim_output.csv'

Returns

Name	Type	Description
	pd.DataFrame	pd.DataFrame: Final dataset containing all successful runs.

Examples

# 1. Define bounds
ranges = {'Length': (0, 10), 'Angle': (-45, 45)}

# 2. Define a command that reads {input} and writes {output}
cmd = (
    "python -c "
    "'import pandas as pd; "
    'df=pd.read_csv("{input}"); '
    'df["Signal"] = df["Length"]*2; '
    'df.to_csv("{output}", index=False)'
    "'"
)

# 3. Run with a 1.5 hour time limit
final_df = run_adaptive_search(
    command=cmd,
    input_cols=['Length', 'Angle'],
    outcome_col='Signal',
    ranges=ranges,
    max_iter=10,
    max_hours=1.5
)
print(f"Collected {len(final_df)} samples.")