写一个xgboost预测股价是否涨跌超过1%的模型，结果有不超过，涨大于1%，跌大约1%

要创建一个基于XGBoost的模型来预测股票价格涨跌是否超过1%，你需要先准备数据、选择特征、训练模型并进行预测。以下是一个简化的Python示例，使用pandas和xgboost库：

# 导入所需库
import pandas as pd
import numpy as np
from sklearn.model_selection import train_test_split
import xgboost as xgb
from sklearn.preprocessing import LabelEncoder

# 假设你有一个包含历史股价数据的数据frame 'stock_data'
data = ... # 股票数据清洗后格式化

# 准备特征（假设我们选择收盘价、开盘价、最高价、最低价等作为特征）
features = ['Close', 'Open', 'High', 'Low']
target = 'ChangePercentage'

# 数据预处理
stock_data['ChangePercentage'] = stock_data[target].apply(lambda x: abs(x) / 100 - 1)  # 计算涨跌幅百分比
stock_data['direction'] = np.where(stock_data['ChangePercentage'] > 0.01, 'Up', 'Down')  # 标记涨跌

# 将标签转换为数字编码
le = LabelEncoder()
stock_data['direction_encoded'] = le.fit_transform(stock_data['direction'])

# 划分训练集和测试集
X_train, X_test, y_train, y_test = train_test_split(stock_data[features], stock_data['direction_encoded'], test_size=0.2, random_state=42)

# 创建DMatrix格式数据
dtrain = xgb.DMatrix(X_train, label=y_train)
dtest = xgb.DMatrix(X_test, label=y_test)

# 设置XGBoost模型参数
params = {
    'objective': 'binary:logistic',  # 使用二分类模型
    'eval_metric': 'logloss',  # 评估指标
    'eta': 0.1,  # 学习率
    'max_depth': 5,  # 深度
}

# 训练模型
model = xgb.train(params, dtrain, num_boost_round=100, evals=[(dtest, 'test')])

# 预测新数据
new_data = ...  # 新一天的股票数据
prediction_df = pd.DataFrame(new_data[features])
prediction_df['ChangePercentage'] = new_data[target]
prediction_df['direction_encoded'] = model.predict(xgb.DMatrix(prediction_df[features]))

# 解码预测结果
prediction_df['direction'] = prediction_df['direction_encoded'].map(le.inverse_transform)