sklearn随机搜索出现问题

问题遇到的现象和发生背景

使用sklearn和lightGBM进行训练，在进行随机搜索时，出现问题：

问题相关代码，请勿粘贴截图

import random
from sklearn.model_selection import RandomizedSearchCV
from sklearn.metrics import accuracy_score
import lightgbm as lgb
import numpy as np
import sys
import os
from osgeo import gdal
import joblib

# 1. 随机选取几块进行数据集创建

res = []  # 存放数据标签，list，str
# 在标签中选出
for i in range(0, 7):
    for j in range(0, 16):
        res.append(str(i) + '_' + str(j))

def sample_block_to_name(seq, block):
    """
    随机选取blck个数，并返回对应的标签名称
    :param seq: 输入初始标签数量
    :param block: 定义随机选取的数
    :return: 标签名称和新的剩余数量
    """
    num = random.sample(seq, block)
    name = []
    for k in num:
        name.append(res[k])
        seq.remove(k)
    return name, seq

seq_original = list(range(0, 112))

name_original = []
for i in range(0, 7):
    for j in range(0, 16):
        name_original.append(str(i) + '_' + str(j))

name, seq_new = sample_block_to_name(seq_original, 70)
print(name)
print(seq_new)
# 读取tif文件, 后续tif_to_array中使用
def Read_img2array(img_file_path):
    """
    读取栅格数据，将其转换成对应数组
    img_file_path: 栅格数据路径
    :return: 返回投影，几何信息，和转换后的数组
    """
    dataset = gdal.Open(img_file_path)  # 读取栅格数据
    # 判断是否读取到数据
    if dataset is None:
        print('Unable to open *.tif')
        sys.exit(1)  # 退出

    # 直接读取dataset
    img_array = dataset.ReadAsArray()
    return img_array

# 将选取的特征区块tif转换成array
def tif_to_array(path, name):
    """
    将选取的所有特征区块转变为数组
    :param path: 文件路径，不到.tif文件
    :param name: 区块
    :return: 返回数组
    """
    pathDir = os.listdir(path)
    result = [[0] * 15] * 1
    for x in name:
        for y in pathDir:  # 0_1,0_1...
            if x == y:
                pathDir_y = os.listdir(path + "/" + y)
                arr = [[0] * 1] * 41 * 36
                arr = np.array(arr)
                arr = arr.reshape(-2, 1)
                for i in pathDir_y:
                    index = i.rfind('.')
                    if i[index:] == '.tif':
                        tif = Read_img2array(path + "/" + y + "/" + i)
                        mul = np.array(tif).reshape(-2, 1)
                        arr = np.column_stack((arr, mul))
                arr = arr[:, 1:]  # 第一个区块的数组
                result = np.concatenate((result, arr), axis=0)
    result = result[1:, :]
    return result

# 将选取的标签区块tif转换成array
def tif_to_array_label(path, name):
    pathdir = os.listdir(path)
    rest = [[0] * 1] * 1
    for x in name:
        for y in pathdir:
            if y == 'dilei' + x + '.tif':
                tif = Read_img2array(path + '/' + y)
                tif = tif.reshape(-2, 1)
                rest = np.concatenate((rest, tif), axis=0)
    rest = rest[1:, :].ravel()
    return rest


####################################################
iteration = 0

while iteration <= 5:
    # 数据、标签
    X_train = tif_to_array("D:\\Personality\\paper\\GBDT\\testshiyan\\train", name)
    y_train = tif_to_array_label("D:\\Personality\\paper\\GBDT\\testshiyan\\label", name)

    X_test = tif_to_array("D:\\Personality\\paper\\GBDT\\testshiyan\\train", name_original)
    y_test = tif_to_array_label("D:\\Personality\\paper\\GBDT\\testshiyan\\label", name_original)

    # 2.随机搜索寻找模型最优参数
    clf = lgb.LGBMClassifier()  # LightGBM参数默认
    param_test = {'n_estimators': list(range(200, 1000, 100)), 'num_leaves': list(range(10, 40, 2)),
                  'learning_rate': list(np.arange(0, 1, 0.1)), 'max_depth': list(range(-1, 50, 1))}  # 设置搜索范围
    rand = RandomizedSearchCV(clf, param_test, scoring='accuracy', n_iter=10,
                              random_state=5, n_jobs=1)  # 随机搜索，默认cv为5折
    rand.fit(X_train, y_train)  # 拟合
    print('随机搜索-最佳参数：', rand.best_params_)  # 获取最佳度量值时的代定参数的值。是一个字典
    print('随机搜索-score-train：', rand.score(X_train, y_train))
    print('随机搜索-score-test：', rand.score(X_test, y_test))

    # 3.训练模型
    params = dict(rand.best_params_)  # 转化为字典
    setting = {"boosting_type": 'gbdt', "subsample": 1.0}
    params.update(setting)  # 更新字典键值对
    clf_new = lgb.LGBMClassifier(**params)  # 训练模型
    clf_new.fit(X_train, y_train)  # 拟合

    # 4.预测，存储预测结果
    y_gbr = clf_new.predict(X_test)  # 测试集进行预测
    accuracy = accuracy_score(y_gbr, y_test)  # 计算准确率
    print("精确率为：")
    print(accuracy)

    # 5.查看预测较低的块
    n = 41 * 36
    acc = []
    for i in range(0, 165312, n):
        y_test1 = y_test[i:i + n]
        y_gbr1 = y_gbr[i:i + n]
        acc1 = accuracy_score(y_gbr1, y_test1)
        acc.append(acc1)
    sort_new = np.argsort(acc)
    sor_new = []
    for i in sort_new:
        for j in seq_new:
            if i == j:
                sor_new.append(i)

    # 6.添加随机块数据集
    for i in sor_new[0:5]:
        name.append(res[i])
        seq_new.remove(i)

    joblib.dump(clf_new, f'train_model_result{iteration}.m')  # 保存模型
    iteration += 1

运行结果及报错内容

我的解答思路和尝试过的方法

我想要达到的结果

score= ' raise '