feat(sampler): add --ratios

icfree/sampler.py

@@ -5,14 +5,15 @@
 from pyDOE2 import lhs
 import ast
 
-def generate_lhs_samples(input_file, num_samples, step, fixed_values=None, seed=None):
+def generate_lhs_samples(input_file, num_samples, step=None, ratios=None, fixed_values=None, seed=None):
     """
     Generates Latin Hypercube Samples for components based on discrete ranges.
     
     Parameters:
     - input_file: Path to the input file containing components and their max values.
     - num_samples: Number of samples to generate.
     - step: Step size for creating discrete ranges.
+    - ratios: List of ratios for creating discrete ranges.
     - fixed_values: Dictionary of components with fixed values (optional).
     - seed: Random seed for reproducibility.
     
@@ -33,11 +34,17 @@ def generate_lhs_samples(input_file, num_samples, step, fixed_values=None, seed=
     # Generate discrete ranges for each component
     for index, row in components_df.iterrows():
         component_name = row['Component']
+        max_value = row['maxValue']
         if fixed_values and component_name in fixed_values:
             # If the component has a fixed value, use a single-element array
             component_range = np.array([fixed_values[component_name]])
         else:
-            component_range = np.arange(0, row['maxValue'] + step, step)
+            if ratios is not None:
+                # Use ratios to create the discrete range
+                component_range = np.array([r * max_value for r in ratios])
+            else:
+                # Use step to create the discrete range
+                component_range = np.arange(0, max_value + step, step)
         discrete_ranges.append(component_range)
 
     # Determine the number of components
@@ -55,15 +62,16 @@ def generate_lhs_samples(input_file, num_samples, step, fixed_values=None, seed=
     samples_df = pd.DataFrame(samples, columns=components_df['Component'])
     return samples_df
 
-def main(input_file, output_file, num_samples, step=2.5, fixed_values=None, seed=None):
+def main(input_file, output_file, num_samples, step=None, ratios=None, fixed_values=None, seed=None):
     """
     Main function to generate LHS samples and save them to a CSV file.
     
     Parameters:
     - input_file: Path to the input file containing components and their max values.
     - output_file: Path to the output CSV file where samples will be written.
     - num_samples: Number of samples to generate.
-    - step: Step size for creating discrete ranges (default: 2.5).
+    - step: Step size for creating discrete ranges (optional).
+    - ratios: List of ratios for creating discrete ranges (optional).
     - fixed_values: Dictionary of components with fixed values (optional).
     - seed: Random seed for reproducibility (optional).
     """
@@ -80,7 +88,7 @@ def main(input_file, output_file, num_samples, step=2.5, fixed_values=None, seed
                 print(f"Warning: Component '{component}' not found in the input file.")
 
     # Generate LHS samples
-    samples_df = generate_lhs_samples(input_file, num_samples, step, fixed_values, seed)
+    samples_df = generate_lhs_samples(input_file, num_samples, step, ratios, fixed_values, seed)
 
     # Write the samples to a CSV file
     samples_df.to_csv(output_file, index=False)
@@ -92,7 +100,12 @@ def main(input_file, output_file, num_samples, step=2.5, fixed_values=None, seed
     parser.add_argument('input_file', type=str, help='Input file path with components and their max values.')
     parser.add_argument('output_file', type=str, help='Output CSV file path for the samples.')
     parser.add_argument('num_samples', type=int, help='Number of samples to generate.')
-    parser.add_argument('--step', type=float, default=2.5, help='Step size for creating discrete ranges (default: 2.5).')
+
+    # Create a mutually exclusive group for step and ratios
+    group = parser.add_mutually_exclusive_group(required=True)
+    group.add_argument('--step', type=float, help='Step size for creating discrete ranges.')
+    group.add_argument('--ratios', type=str, help='Comma-separated list of ratios for creating discrete ranges (e.g., "0,0.2,0.4,0.6,0.8,1").')
+
     parser.add_argument('--fixed_values', type=str, default=None, help='Fixed values for components as a dictionary (e.g., \'{"Component1": 10, "Component2": 20}\')')
     parser.add_argument('--seed', type=int, default=None, help='Seed for random number generation for reproducibility (optional).')
 
@@ -102,5 +115,8 @@ def main(input_file, output_file, num_samples, step=2.5, fixed_values=None, seed
     # Convert fixed_values argument from string to dictionary if provided
     fixed_values = ast.literal_eval(args.fixed_values) if args.fixed_values else None
 
+    # Convert ratios argument from comma-separated string to list of floats if provided
+    ratios = [float(r) for r in args.ratios.split(',')] if args.ratios else None
+
     # Run the main function with the parsed arguments
-    main(args.input_file, args.output_file, args.num_samples, args.step, fixed_values, args.seed)
+    main(args.input_file, args.output_file, args.num_samples, args.step, ratios, fixed_values, args.seed)