Genome

Overview

The Genome class defines the complete genetic representation of a trading bot’s strategy. All gene values are stored as floats in the range [0.0, 1.0] and decoded into actual trading parameters by the decode_genome() function.

Genome Class

Fields

Identity

string

Unique 8-character identifier (auto-generated)

generation

int

default:"0"

Generation number this genome belongs to

parent_ids

list[str]

default:"[]"

IDs of parent genomes (1 for clone/elite, 2 for crossover)

Market Selection Genes (5 genes)

min_volume_24h

float

default:"0.5"

Minimum 24h volume filter (decoded: 0-5000)

min_open_interest

float

default:"0.5"

Minimum open interest filter (decoded: 0-2000)

min_time_to_expiry_hrs

float

default:"0.3"

Minimum hours until market expiry (decoded: 0-24h)

max_time_to_expiry_hrs

float

default:"0.7"

Maximum hours until market expiry (decoded: 0-24h)

category_mask

float

default:"0.5"

Bitmask for market categories to trade (decoded to category list)

Entry Signal Genes (8 genes)

signal_type

float

default:"0.5"

Signal strategy selector (decoded: price_level, momentum, mean_reversion, value, contrarian)

price_threshold_low

float

default:"0.3"

Low price threshold (decoded: 0.01-0.50)

price_threshold_high

float

default:"0.7"

High price threshold (decoded: 0.50-0.99)

momentum_lookback

float

default:"0.5"

Momentum lookback period (decoded: 1-60 ticks)

momentum_trigger

float

default:"0.5"

Momentum trigger threshold (decoded: -10% to +10%)

mean_rev_zscore

float

default:"0.5"

Mean reversion z-score threshold (decoded: 0.5-3.0)

value_edge_min

float

default:"0.5"

Minimum value edge required (decoded: 0.01-0.30)

contrarian_threshold

float

default:"0.5"

Contrarian confidence threshold (decoded: 0.60-0.95)

Side Selection Genes (2 genes)

side_bias

float

default:"0.5"

Directional bias: less than 0.2 = always no, 0.2-0.8 = signal, greater than 0.8 = always yes

side_flip_prob

float

default:"0.0"

Probability of randomly flipping side (decoded: 0-0.5)

Position Sizing Genes (3 genes)

bankroll_fraction

float

default:"0.02"

Fraction of bankroll to risk per trade (decoded: 0.005-0.10)

max_concurrent_positions

float

default:"0.3"

Maximum concurrent open positions (decoded: 1-20)

max_single_market_pct

float

default:"0.5"

Max allocation to single market (decoded: 0.01-0.25)

Risk Management Genes (4 genes)

daily_loss_limit_pct

float

default:"0.5"

Daily loss limit as % of equity (decoded: 0.02-0.30)

max_trades_per_day

float

default:"0.5"

Maximum trades per day (decoded: 1-100)

min_price

float

default:"0.1"

Minimum acceptable contract price (decoded: 0.01-0.50)

max_price

float

default:"0.9"

Maximum acceptable contract price (decoded: 0.50-0.99)

Class Methods

gene_names()

Get list of all evolvable gene field names.

genes

list[str]

List of 22 gene field names

from genetic.genome import Genome

genes = Genome.gene_names()
print(f"Genome has {len(genes)} genes: {genes}")
# Output: Genome has 22 genes: ['min_volume_24h', 'min_open_interest', ...]

random(generation=0)

Create a genome with all genes randomized uniformly in [0, 1].

generation

int

default:"0"

Generation number to assign

genome

Genome

New genome with random gene values

# Create random genome for generation 5
genome = Genome.random(generation=5)
print(f"ID: {genome.id}")
print(f"Signal type gene: {genome.signal_type:.3f}")
print(f"Bankroll fraction: {genome.bankroll_fraction:.3f}")

Instance Methods

clone()

Create a deep copy with a new unique ID.

genome

Genome

Cloned genome with new ID

original = Genome.random()
clone = original.clone()

print(f"Original ID: {original.id}")
print(f"Clone ID: {clone.id}")
print(f"Same genes: {original.signal_type == clone.signal_type}")

to_dict()

Serialize genome to dictionary for JSON persistence.

data

dict

Dict containing all genome fields

import json

genome = Genome.random()
data = genome.to_dict()

# Save to file
with open("genome.json", "w") as f:
    json.dump(data, f, indent=2)

from_dict(d)

Deserialize genome from dictionary.

dict

required

Dictionary containing genome fields

genome

Genome

Reconstructed genome instance

import json

# Load from file
with open("genome.json") as f:
    data = json.load(f)

genome = Genome.from_dict(data)
print(f"Loaded genome {genome.id} from generation {genome.generation}")

Signal Types

The signal_type gene is discretized into one of these strategies:

SIGNAL_TYPES = [
    "price_level",      # Buy when price in specific range
    "momentum",         # Buy based on price direction
    "mean_reversion",   # Buy on z-score deviation
    "value",            # Buy cheap side vs 50/50 fair value
    "contrarian"        # Bet against crowd when confident
]

decode_genome Function

Convert [0,1] gene values into actual trading parameters.

genome

Genome

required

Genome to decode

known_categories

list[str]

required

List of market categories available

params

dict

Decoded trading parameters ready for bot execution

from genetic.genome import Genome, decode_genome

genome = Genome.random()
categories = ["Finance", "Politics", "Economics", "Science"]

params = decode_genome(genome, categories)

print(f"Signal type: {params['signal_type']}")
print(f"Bankroll fraction: {params['bankroll_fraction']:.2%}")
print(f"Max concurrent: {params['max_concurrent']}")
print(f"Min volume: ${params['min_volume_24h']:,.0f}")
print(f"Trading categories: {params['categories']}")
print(f"Price range: ${params['min_price']:.2f}-${params['max_price']:.2f}")

Example decoded output:

{
    # Market selection
    "min_volume_24h": 2500.0,
    "min_open_interest": 1000.0,
    "min_time_to_expiry_hrs": 4.5,
    "max_time_to_expiry_hrs": 18.2,
    "categories": ["Finance", "Economics"],
    
    # Entry signal
    "signal_type": "momentum",
    "price_threshold_low": 0.35,
    "price_threshold_high": 0.72,
    "momentum_lookback_ticks": 30,
    "momentum_trigger_pct": 0.05,
    "mean_rev_zscore": 2.0,
    "value_edge_min": 0.15,
    "contrarian_threshold": 0.80,
    
    # Side selection
    "side_bias": 0.5,
    "side_flip_prob": 0.15,
    
    # Position sizing
    "bankroll_fraction": 0.025,
    "max_concurrent": 8,
    "max_single_market_pct": 0.12,
    
    # Risk management
    "daily_loss_limit_pct": 0.15,
    "max_trades_per_day": 50,
    "min_price": 0.20,
    "max_price": 0.75
}

Gene Evolution

Genes evolve through:

Crossover: Random mix of parent genes
Mutation: Gaussian perturbation clamped to [0,1]
Immigration: Fresh random genomes for diversity

See the Evolution reference for details.

Core Modules

Genetic Modules

Configuration

Overview

Genome Class

Fields

Identity

Market Selection Genes (5 genes)

Entry Signal Genes (8 genes)

Side Selection Genes (2 genes)

Position Sizing Genes (3 genes)

Risk Management Genes (4 genes)

Class Methods

gene_names()

random(generation=0)

Instance Methods

clone()

to_dict()

from_dict(d)

Signal Types

decode_genome Function

Gene Evolution

Build docs developers (and LLMs) love

Core Modules

Genetic Modules

Configuration

​Overview

​Genome Class

​Fields

​Identity

​Market Selection Genes (5 genes)

​Entry Signal Genes (8 genes)

​Side Selection Genes (2 genes)

​Position Sizing Genes (3 genes)

​Risk Management Genes (4 genes)

​Class Methods

​gene_names()

​random(generation=0)

​Instance Methods

​clone()

​to_dict()

​from_dict(d)

​Signal Types

​decode_genome Function

​Gene Evolution

Build docs developers (and LLMs) love

Overview

Genome Class

Fields

Identity

Market Selection Genes (5 genes)

Entry Signal Genes (8 genes)

Side Selection Genes (2 genes)

Position Sizing Genes (3 genes)

Risk Management Genes (4 genes)

Class Methods

gene_names()

random(generation=0)

Instance Methods

clone()

to_dict()

from_dict(d)

Signal Types

decode_genome Function

Gene Evolution