Skip to content

config.md

Latest commit

 

History

History
143 lines (109 loc) · 3.29 KB

config.md

File metadata and controls

143 lines (109 loc) · 3.29 KB

Configuration Module Documentation

Overview

The configuration module provides global settings management for MOBPY through the MOBPYConfig dataclass. It supports runtime configuration, environment variables, and persistence.

Module Location

src/MOBPY/config.py

MOBPYConfig Class

Default Configuration

@dataclass
class MOBPYConfig:
    epsilon: float = 1e-12          # Numerical tolerance
    max_iterations: int = 1000      # Max iterations for algorithms
    enable_progress_bar: bool = False  # Progress tracking
    n_jobs: int = 1                 # Parallel processing
    random_state: Optional[int] = None  # RNG seed
    warn_on_constraint_violation: bool = True
    validate_inputs: bool = True    # Input validation
    cache_size: int = 128           # MB for caching
    precision: int = 6              # Decimal precision

Global Functions

get_config()

Gets the global MOBPY configuration singleton.

from MOBPY.config import get_config

config = get_config()
print(f"Current epsilon: {config.epsilon}")

set_config(**kwargs)

Updates global configuration parameters.

from MOBPY.config import set_config

set_config(epsilon=1e-10, n_jobs=4, enable_progress_bar=True)

reset_config()

Resets global configuration to defaults.

from MOBPY.config import reset_config

reset_config()  # Back to all defaults

Class Methods

Instance Management

# Create custom configuration
config = MOBPYConfig(epsilon=1e-10, n_jobs=-1)

# Save to file
config.save("my_config.json")

# Load from file
config = MOBPYConfig.from_file("my_config.json")

# Load from environment variables (MOBPY_ prefix)
config = MOBPYConfig.from_env()

# Export as dictionary
config_dict = config.to_dict()

Environment Variables

Set configuration via environment:

export MOBPY_EPSILON=1e-10
export MOBPY_N_JOBS=4
export MOBPY_ENABLE_PROGRESS_BAR=true

Then load in Python:

config = MOBPYConfig.from_env()

Usage Examples

Basic Configuration

from MOBPY.config import set_config

# Configure for production
set_config(
    epsilon=1e-10,      # Higher precision
    n_jobs=-1,          # Use all cores
    validate_inputs=True,  # Ensure data quality
    enable_progress_bar=False  # No output in production
)

Development Configuration

# Configure for debugging
set_config(
    enable_progress_bar=True,
    warn_on_constraint_violation=True,
    validate_inputs=True,
    precision=8  # More decimal places for debugging
)

Configuration Context Manager

from MOBPY.config import get_config, set_config

# Temporarily change configuration
original_epsilon = get_config().epsilon
try:
    set_config(epsilon=1e-15)
    # Do precision-critical work
finally:
    set_config(epsilon=original_epsilon)

Performance Impact

Setting Impact Recommendation
epsilon Affects convergence Use 1e-12 for most cases
n_jobs Parallel processing -1 for all cores
validate_inputs Adds overhead Disable in production if data is trusted
cache_size Memory usage Increase for large datasets

See Also