使用 Python 自动提交 Numerai 的加密信号

自动提交 Numerai 加密信号

该脚本使用 Python 进行 API 交互，以向 Numerai 的加密信号锦标赛提交预测。该代码侧重于错误处理、模块化和验证。

import pandas as pd
from numerapi import SignalsAPI
import sys
import os
from uuid import uuid4
# Function to load and validate predictions
def load_predictions(file_path):
    if not os.path.exists(file_path):
        raise FileNotFoundError(f"File not found: {file_path}")
    try:
        predictions = pd.read_csv(file_path)
        if "prediction" not in predictions.columns:
            raise ValueError("File must contain a 'prediction' column.")
        return predictions
    except Exception as e:
        raise ValueError(f"Error reading the file: {e}")
# Function to upload predictions
def upload_predictions(api_key, model_id, file_path):
    try:
        api = SignalsAPI(api_key)
        api.upload_predictions(file_path, model_id=model_id)
        print(f"Predictions uploaded successfully for model ID: {model_id}")
    except Exception as e:
        print(f"Failed to upload predictions: {e}")
# Main execution
if __name__ == "__main__":
    if len(sys.argv) != 3:
        print("Usage: python submit_signals.py <api_key> <predictions_file_path>")
        sys.exit(1)
    api_key = sys.argv[1]
    predictions_file_path = sys.argv[2]
    model_id = str(uuid4())  # Replace with actual model ID
    try:
        load_predictions(predictions_file_path)
        upload_predictions(api_key, model_id, predictions_file_path)
    except Exception as e:
        print(f"An error occurred: {e}")

基于 CLI 的 Numerai 加密信号提交

此示例利用 Numerai 的 CLI 进行提交，为熟悉终端命令的用户提供更简单的方法。

#!/bin/bash
# Numerai CLI submission script
# Validate input arguments
if [ "$#" -ne 3 ]; then
    echo "Usage: ./submit.sh <model_id> <api_key> <predictions_file_path>"
    exit 1
fi
MODEL_ID=$1
API_KEY=$2
PREDICTIONS_FILE=$3
# Check if file exists
if [ ! -f "$PREDICTIONS_FILE" ]; then
    echo "Error: File $PREDICTIONS_FILE does not exist."
    exit 1
fi
# Execute Numerai CLI submission
numerai signals upload --model-id "$MODEL_ID" --apikey "$API_KEY" --file "$PREDICTIONS_FILE"
if [ $? -eq 0 ]; then
    echo "Predictions submitted successfully for Model ID: $MODEL_ID"
else
    echo "Submission failed. Check your inputs and try again."
fi

对 Python 解决方案进行单元测试

本部分包括一个 Python 单元测试脚本，用于验证所提供的 Python 解决方案的功能。

import unittest
from unittest.mock import patch
import os
from your_script import load_predictions, upload_predictions
class TestNumeraiSubmission(unittest.TestCase):
    def test_load_predictions_valid(self):
        file_path = "valid_predictions.csv"
        pd.DataFrame({"prediction": [0.1, 0.2]}).to_csv(file_path, index=False)
        try:
            predictions = load_predictions(file_path)
            self.assertIn("prediction", predictions.columns)
        finally:
            os.remove(file_path)
    def test_load_predictions_missing_file(self):
        with self.assertRaises(FileNotFoundError):
            load_predictions("missing_file.csv")
    @patch("your_script.SignalsAPI")
    def test_upload_predictions_success(self, mock_api):
        mock_instance = mock_api.return_value
        mock_instance.upload_predictions.return_value = None
        upload_predictions("dummy_key", "dummy_model", "dummy_path")
        mock_instance.upload_predictions.assert_called_once()
if __name__ == "__main__":
    unittest.main()

1. 官方 Numerai Signals API 文档：有关 API 函数的详细信息以及提交预测的示例。 Numerai 信号 API
2. Pandas 库文档：有关使用 Pandas 进行数据操作和验证的综合指南。 熊猫文档
3. Python 单元测试文档：有关设置和运行 Python 脚本单元测试的说明。 Python 单元测试
4. Numerai CLI 指南：通过命令行提交预测的步骤。 Numerai CLI GitHub
5. Python os 模块文档：有关在 Python 中管理文件路径和验证文件存在的信息。 Python 操作系统模块