Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/DeweyMarco/simple-kalshi-bot/llms.txt

Use this file to discover all available pages before exploring further.

​
Overview

The monitor.py module provides logging setup and progress monitoring functions for the genetic algorithm evolution loop. It handles dual logging (file + console), periodic progress reports, and generation summaries with leaderboards.

​
Functions

​
setup_logging() → logging.Logger

Configure dual logging with file and console handlers.
​
logger
logging.Logger
Configured logger instance named “evolution”
Behavior:
  • Creates STATE_DIR/evolution.log for full debug logs
  • Outputs INFO+ messages to console with timestamps
  • Automatically creates state directory if missing
  • Prevents duplicate handlers on re-initialization
from genetic.monitor import setup_logging

logger = setup_logging()
logger.info("Evolution started")
logger.debug("Detailed debug info (file only)")
Log Formats:
  • File: 2023-12-31 10:30:45 [INFO] message
  • Console: [10:30:45] message

​
log_tick_progress()

Log periodic progress during a generation’s trading period.
​
logger
logging.Logger
required
Logger instance from setup_logging()
​
gen_num
int
required
Current generation number
​
tick_count
int
required
Number of ticks elapsed in this generation
​
bots
list[GeneticBot]
required
List of all bots in this generation
​
feed
MarketDataFeed | None
Optional market data feed for unrealized ROI calculation
Output Example: 
[Gen 5 | 2.5h] Active: 45/50 | Open: 120 | Trades: 458 | Settled: 203 | 
Est ROI: +15.2%/+8.1% | Realized: +12.3%/+6.5%

from genetic.monitor import log_tick_progress

for tick in range(generation_ticks):
    # ... trading logic ...
    
    if tick % PROGRESS_LOG_INTERVAL_TICKS == 0:
        log_tick_progress(logger, gen_num, tick, bots, feed)
Metrics Reported:
  • Elapsed time in hours
  • Active bots (with at least one trade)
  • Total open positions across all bots
  • Total trades and settled positions
  • Best and median ROI (both estimated and realized)

​
log_generation_summary()

Log a complete generation summary with leaderboard after trading completes.
​
logger
logging.Logger
required
Logger instance from setup_logging()
​
gen_num
int
required
Generation number being summarized
​
bots
list[GeneticBot]
required
List of all bots in this generation
​
fitness
list[float]
required
Fitness scores (ROI percentages) for each bot
Output Example: 
======================================================================
GENERATION 5 RESULTS
======================================================================
Rank  Bot ID         ROI%    Trades Settled  WinRate          Signal
----------------------------------------------------------------------
1     a1b2c3d4       18.5%      127      89    65.2%     momentum
2     e5f6g7h8       15.2%       98      67    58.2%  price_level
3     i9j0k1l2       12.8%      115      78    61.5%         value
...

Signal distribution: {'momentum': 15, 'price_level': 12, 'value': 10, ...}
Top-10 category prefs: {'politics': 8, 'economics': 6, 'crypto': 4, ...}

from genetic.monitor import log_generation_summary
from genetic.evolution import evaluate_fitness

# After generation completes
fitness = [evaluate_fitness(bot) for bot in bots]
log_generation_summary(logger, gen_num, bots, fitness)
Information Displayed:
  • Top 20 bots ranked by fitness (ROI)
  • Bot ID, ROI%, total trades, settled trades, win rate, signal type
  • Distribution of signal types across population
  • Category preferences of top 10 performers

​
compute_generation_stats(bots: list[GeneticBot]) → dict

Compute summary statistics for a completed generation.
​
bots
list[GeneticBot]
required
List of all bots in the generation
​
stats
dict
Dictionary containing:
  • best_roi: Highest fitness (ROI%) in generation
  • median_roi: Median fitness across all bots
  • worst_roi: Lowest fitness in generation
  • mean_roi: Average fitness across all bots
  • mean_trades: Average number of trades per bot
  • max_trades: Maximum trades by any single bot
  • mean_win_rate: Average win rate percentage
  • total_settled: Total settled positions across all bots
  • active_bots: Number of bots that made at least one trade
from genetic.monitor import compute_generation_stats

stats = compute_generation_stats(bots)
print(f"Best ROI: {stats['best_roi']:.2f}%")
print(f"Median ROI: {stats['median_roi']:.2f}%")
print(f"Mean trades: {stats['mean_trades']:.1f}")
print(f"Active bots: {stats['active_bots']}/{len(bots)}")

​
Usage Example

Complete monitoring setup for an evolution loop: 
from genetic.monitor import (
    setup_logging,
    log_tick_progress,
    log_generation_summary,
    compute_generation_stats
)
from genetic.config import PROGRESS_LOG_INTERVAL_TICKS

# Initialize
logger = setup_logging()
logger.info("Starting evolution...")

# During generation
for tick in range(total_ticks):
    # Trading logic here
    for bot in bots:
        bot.tick()
    
    # Periodic progress
    if tick % PROGRESS_LOG_INTERVAL_TICKS == 0:
        log_tick_progress(logger, gen_num, tick, bots, feed)

# After generation
fitness = [evaluate_fitness(bot) for bot in bots]
log_generation_summary(logger, gen_num, bots, fitness)

stats = compute_generation_stats(bots)
logger.info(f"Best: {stats['best_roi']:.2f}%, Active: {stats['active_bots']}")

​
Configuration

Uses these constants from genetic.config:
  • STATE_DIR: Directory for log file storage
  • TICK_INTERVAL_SECONDS: Used to calculate elapsed hours

Build docs developers (and LLMs) love

Get started for freeTalk to us