使用PyInstaller打包含modin的Python脚本为exe反复缺依赖如何解决

你遇到的报错本质是两类PyInstaller打包常见问题：

• 动态导入的依赖模块：modin、swifter、dask/distributed这类数据处理库大量使用延迟导入/动态导入逻辑，PyInstaller的静态依赖扫描无法识别到这些隐式依赖，导致运行时缺模块
• 非py格式资源文件：distributed.yaml这类库自带的配置文件不属于Python模块，默认不会被打包进exe的运行目录

在你脚本最开头、所有第三方库导入之前，添加以下代码：

import os
# 提前固定modin引擎，避免运行时读取环境变量失败
os.environ["MODIN_ENGINE"] = "dask"
# 可选：关闭swifter的自动Modin fallback，避免索引对齐异常
import swifter
swifter.set_dask_scheduler("synchronous")

如果遇到ValueError: Can only compare identically-labeled Series objects报错，在调用apply的DataFrame上先执行df = df.reset_index(drop=True)统一索引即可。 #

第一步 生成基础spec文件 #

在你的虚拟环境中执行命令生成初始spec：

pyi-makespec DTC_dask.py

第二步 编辑spec文件 #

替换生成的spec文件内容如下，注意修改site_packages_path为你当前虚拟环境的site-packages目录路径（可执行python -c "import site; print(site.getsitepackages()[0])"获取路径）：

# -*- mode: python ; coding: utf-8 -*-
import os
# 替换为你虚拟环境的site-packages实际路径
site_packages_path = r"C:\Users\TTC01017\Envs\your_env\Lib\site-packages"

block_cipher = None

a = Analysis(
    ['DTC_dask.py'],
    pathex=[],
    binaries=[],
    # 配置资源文件：把distributed、modin的配置文件打包到对应目录
    datas=[
        (os.path.join(site_packages_path, "distributed", "*.yaml"), "distributed"),
        (os.path.join(site_packages_path, "distributed", "*.yml"), "distributed"),
        (os.path.join(site_packages_path, "modin", "config"), "modin/config"),
        (os.path.join(site_packages_path, "swifter"), "swifter"),
    ],
    # 配置隐藏导入模块，覆盖所有动态导入的依赖
    hiddenimports=[
        # modin相关依赖
        "modin", "modin.config", "modin.pandas", "modin.core.execution.dask",
        "modin.core.execution.ray", "modin.config.envvars", "modin.config.pubsub",
        # distributed/dask相关依赖
        "distributed", "distributed.config", "dask", "dask.config", "cloudpickle",
        "tornado", "tblib", "zict", "psutil", "click", "msgpack",
        # swifter相关依赖
        "swifter", "swifter.swifter",
        # pandas隐藏依赖
        "pandas._libs.tslibs.timedeltas", "pandas._libs.tslibs.np_datetime",
        "pandas._libs.tslibs.conversion", "pandas._libs.algos"
    ],
    hookspath=[],
    hooksconfig={},
    runtime_hooks=[],
    excludes=[],
    win_no_prefer_redirects=False,
    win_private_assemblies=False,
    cipher=block_cipher,
    noarchive=False,
)
pyz = PYZ(a.pure, a.zipped_data, cipher=block_cipher)

exe = EXE(
    pyz,
    a.scripts,
    a.binaries,
    a.zipfiles,
    a.datas,
    [],
    name='DTC_dask',
    debug=False,
    bootloader_ignore_signals=False,
    strip=False,
    upx=True,
    upx_exclude=[],
    runtime_tmpdir=None,
    console=True,
    disable_windowed_traceback=False,
    argv_emulation=False,
    target_arch=None,
    codesign_identity=None,
    entitlements_file=None,
)

第三步 执行打包 #

执行以下命令用spec文件打包：

pyinstaller DTC_dask.spec

打包完成后可执行文件会生成在dist目录下。 #

如果后续仍然出现ModuleNotFoundError，直接把报错的模块名添加到spec的hiddenimports列表中重新打包即可。

内容的提问来源于stack exchange，提问作者sohail22b