理解如何让库识别我有所需的依赖项

Question

我开始了一个项目，需要使用一个叫做MuJoCo的开源软件，这是MuJoCo的文档。这个软件是用来做物理模拟的，特别适合强化学习（RL）。因为我正在开发一个人工智能应用，所以我选择用Python来编程，这样比较方便。

我遇到的问题是，MuJoCo使用一个叫做'mediapy'的库来显示视频帧序列，而这个库又依赖于另一个叫做'ffmpeg'的库。问题是'mediapy'没有识别到我的电脑上已经安装了'ffmpeg'（我用的是Mac，Intel处理器）。我深入查看了源代码，想找出问题所在，发现'mediapy'是通过运行以下代码来检查'ffmpeg'是否安装的：

import shutil
import typing
import os

_Path = typing.Union[str, 'os.PathLike[str]']

class _Config:
  ffmpeg_name_or_path: _Path = 'ffmpeg'
  show_save_dir: _Path | None = None

_config = _Config()

def _search_for_ffmpeg_path() -> str | None:
  """Returns a path to the ffmpeg program, or None if not found."""
  if filename := shutil.which(_config.ffmpeg_name_or_path):
    return str(filename)
  return None

然后我找到了我所有安装的库所在的文件夹，路径是opt/anaconda3/envs/myenv/lib/python3.10/site-packages，发现里面有一个'ffmpeg'的文件夹，里面的内容和其他库是一样的。所以它显然是存在的。但是当我运行上面的代码时，返回的结果是'None'。我明显漏掉了什么，但就是找不到。

为了参考，这段代码引发了这个问题：

tippe_top = """
<mujoco model="tippe top">
  <option integrator="RK4"/>

  <asset>
    <texture name="grid" type="2d" builtin="checker" rgb1=".1 .2 .3"
     rgb2=".2 .3 .4" width="300" height="300"/>
    <material name="grid" texture="grid" texrepeat="8 8" reflectance=".2"/>
  </asset>

  <worldbody>
    <geom size=".2 .2 .01" type="plane" material="grid"/>
    <light pos="0 0 .6"/>
    <camera name="closeup" pos="0 -.1 .07" xyaxes="1 0 0 0 1 2"/>
    <body name="top" pos="0 0 .02">
      <freejoint/>
      <geom name="ball" type="sphere" size=".02" />
      <geom name="stem" type="cylinder" pos="0 0 .02" size="0.004 .008"/>
      <geom name="ballast" type="box" size=".023 .023 0.005"  pos="0 0 -.015"
       contype="0" conaffinity="0" group="3"/>
    </body>
  </worldbody>

  <keyframe>
    <key name="spinning" qpos="0 0 0.02 1 0 0 0" qvel="0 0 0 0 1 200" />
  </keyframe>
</mujoco>
"""
model = mujoco.MjModel.from_xml_string(tippe_top)
renderer = mujoco.Renderer(model)
data = mujoco.MjData(model)

duration = 7    # (seconds)
framerate = 60  # (Hz)

# Simulate and display video.
frames = []
mujoco.mj_resetDataKeyframe(model, data, 0)  # Reset the state to keyframe 0
while data.time < duration:
  mujoco.mj_step(model, data)
  if len(frames) < data.time * framerate:
    renderer.update_scene(data, "closeup")
    pixels = renderer.render()
    frames.append(pixels)

media.show_video(frames, fps=framerate)

接下来是出现的错误信息：

{
    "name": "RuntimeError",
    "message": "Program 'ffmpeg' is not found; perhaps install ffmpeg using 'apt install ffmpeg'.",
    "stack": "---------------------------------------------------------------------------
RuntimeError                              Traceback (most recent call last)
Cell In[4], line 14
     11     pixels = renderer.render()
     12     frames.append(pixels)
---> 14 media.show_video(frames, fps=framerate)

File ~/opt/anaconda3/envs/Exjobb/lib/python3.10/site-packages/mediapy/__init__.py:1858, in show_video(images, title, **kwargs)
   1835 def show_video(
   1836     images: Iterable[_NDArray], *, title: str | None = None, **kwargs: Any
   1837 ) -> str | None:
   1838   \"\"\"Displays a video in the IPython notebook and optionally saves it to a file.
   1839 
   1840   See `show_videos`.
   (...)
   1856     html string if `return_html` is `True`.
   1857   \"\"\"
-> 1858   return show_videos([images], [title], **kwargs)

File ~/opt/anaconda3/envs/Exjobb/lib/python3.10/site-packages/mediapy/__init__.py:1940, in show_videos(videos, titles, width, height, downsample, columns, fps, bps, qp, codec, ylabel, html_class, return_html, **kwargs)
   1938   video = [resize_image(image, (h, w)) for image in video]
   1939   first_image = video[0]
-> 1940 data = compress_video(
   1941     video, metadata=metadata, fps=fps, bps=bps, qp=qp, codec=codec
   1942 )
   1943 if title is not None and _config.show_save_dir:
   1944   suffix = _filename_suffix_from_codec(codec)

File ~/opt/anaconda3/envs/Exjobb/lib/python3.10/site-packages/mediapy/__init__.py:1777, in compress_video(images, codec, **kwargs)
   1775 with tempfile.TemporaryDirectory() as directory_name:
   1776   tmp_path = pathlib.Path(directory_name) / f'file{suffix}'
-> 1777   write_video(tmp_path, images, codec=codec, **kwargs)
   1778   return tmp_path.read_bytes()

File ~/opt/anaconda3/envs/Exjobb/lib/python3.10/site-packages/mediapy/__init__.py:1747, in write_video(path, images, **kwargs)
   1745   dtype = np.dtype(np.uint16)
   1746 kwargs = {'metadata': getattr(images, 'metadata', None), **kwargs}
-> 1747 with VideoWriter(path, shape=shape, dtype=dtype, **kwargs) as writer:
   1748   for image in images:
   1749     writer.add_image(image)

File ~/opt/anaconda3/envs/Exjobb/lib/python3.10/site-packages/mediapy/__init__.py:1567, in VideoWriter.__enter__(self)
   1566 def __enter__(self) -> 'VideoWriter':
-> 1567   ffmpeg_path = _get_ffmpeg_path()
   1568   input_pix_fmt = self._get_pix_fmt(self.dtype, self.input_format)
   1569   try:

File ~/opt/anaconda3/envs/Exjobb/lib/python3.10/site-packages/mediapy/__init__.py:1167, in _get_ffmpeg_path()
   1165 path = _search_for_ffmpeg_path()
   1166 if not path:
-> 1167   raise RuntimeError(
   1168       f\"Program '{_config.ffmpeg_name_or_path}' is not found;\"
   1169       \" perhaps install ffmpeg using 'apt install ffmpeg'.\"
   1170   )
   1171 return path

RuntimeError: Program 'ffmpeg' is not found; perhaps install ffmpeg using 'apt install ffmpeg'."
}

任何帮助都非常感谢。

我尝试过更改'shutil.which'命令搜索的文件名。我也尝试过重新安装这些库，还试过不同的库（似乎有两个'ffmpeg-python'，一个是0.2版本，另一个是'ffmpeg'，在我的目录中显示为'imageio_ffmpeg'，版本是0.4.9，但在pip网站上显示是1.4）。

开源软件依赖管理错误调试强化学习 FFmpeg 人工智能应用物理模拟 mediapy

理解如何让库识别我有所需的依赖项

1 个回答

撰写回答