关于python:mpld3 3D散点图

mpld3 3D scatter plot

我正在探索MPLD3库,无法确定如何创建三维散点图。有了Matplotlib,我可以:

1
2
3
4
5
6
7
import matplotlib.pyplot as plt
from mpl_toolkits.mplot3d import Axes3D

fig = plt.figure()
ax = fig.add_subplot(111, projection='3d')
ax.scatter([1],[5],[7])
plt.show()

同样,我尝试了MPLD3(在Jupyter笔记本中):将matplotlib.pyplot导入为plt从mpl_toolkits.mplot3d导入axes3d导入MPLD3

1
2
3
4
5
mpld3.enable_notebook()
fig = plt.figure()
ax = fig.add_subplot(111, projection='3d')
ax.scatter([1],[5],[7])
mpld3.display(fig)

我得到了错误

1
TypeError: array([ 0.94,  1.06]) is not JSON serializable

有什么想法吗?

这是完整的错误日志:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
Traceback (most recent call last) <ipython-input-26-54fc4a65da87> in <module>()
      6 N = 50
      7 ax.scatter([1],[5],[7])
----> 8 mpld3.display(fig)

/usr/local/lib/python2.7/dist-packages/mpld3/_display.pyc in display(fig, closefig, local, **kwargs)
    288     if closefig:
    289         plt.close(fig)
--> 290     return HTML(fig_to_html(fig, **kwargs))
    291
    292

/usr/local/lib/python2.7/dist-packages/mpld3/_display.pyc in fig_to_html(fig, d3_url, mpld3_url, no_extras, template_type, figid, use_http, **kwargs)
    234                            d3_url=d3_url,
    235                            mpld3_url=mpld3_url,
--> 236                            figure_json=json.dumps(figure_json),
    237                            extra_css=extra_css,
    238                            extra_js=extra_js)

/usr/lib/python2.7/json/__init__.pyc in dumps(obj, skipkeys, ensure_ascii, check_circular, allow_nan, cls, indent, separators, encoding, default, sort_keys, **kw)
    241         cls is None and indent is None and separators is None and
    242         encoding == 'utf-8' and default is None and not sort_keys and not kw):
--> 243         return _default_encoder.encode(obj)
    244     if cls is None:
    245         cls = JSONEncoder

/usr/lib/python2.7/json/encoder.pyc in encode(self, o)
    205         # exceptions aren't as detailed.  The list call should be roughly
    206         # equivalent to the PySequence_Fast that ''.join() would do.
--> 207         chunks = self.iterencode(o, _one_shot=True)
    208         if not isinstance(chunks, (list, tuple)):
    209             chunks = list(chunks)

/usr/lib/python2.7/json/encoder.pyc in iterencode(self, o, _one_shot)
    268                 self.key_separator, self.item_separator, self.sort_keys,
    269                 self.skipkeys, _one_shot)
--> 270         return _iterencode(o, 0)
    271
    272 def _make_iterencode(markers, _default, _encoder, _indent, _floatstr,

/usr/lib/python2.7/json/encoder.pyc in default(self, o)
    182
    183        """
--> 184         raise TypeError(repr(o) +" is not JSON serializable")
    185
    186     def encode(self, o):

TypeError: array([ 0.94,  1.06]) is not JSON serializable


正如abobe所提到的,mpld3不支持3D图形,但是检查下面的包,它们还打算通过python进行可视化,plotly将matplotlib图形转换为它们的图形,mpld3也支持。

  • https://plot.ly/python/
  • https://bokeh.pydata.org/en/latest/最新/

对不起,这不是一个完整的答案,而是一个很长的评论…

我可以通过编辑mpld3.u display.py来消除错误,如下所示:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
class NumpyEncoder(json.JSONEncoder):
   """ Special json encoder for numpy types"""

    def default(self, obj):
        try:
            obj=obj.tolist()
            return json.JSONEncoder.encode(self, obj)
        except AttributeError:
            pass
        if isinstance(obj, (numpy.int_, numpy.intc, numpy.intp, numpy.int8,
            numpy.int16, numpy.int32, numpy.int64, numpy.uint8,
            numpy.uint16,numpy.uint32, numpy.uint64)):
            return int(obj)
        elif isinstance(obj, (numpy.float_, numpy.float16, numpy.float32,
            numpy.float64)):
            return float(obj)

        return json.JSONEncoder.default(self, obj)

但是,它会在笔记本中生成一个带有错误消息的空绘图:

1
2
3
Javascript error adding output!
TypeError: t.map is not a function
See your browser Javascript console for more details.

看来MPLD3不是为支持3D而设计的…


不幸的是,目前MPLD3不支持3D绘图。它只支持二维绘图。检查以下链接:

  • https://github.com/mpld3/mpld3/issues/223(https://github.com/mpld3/mpld3/issues/223)
  • python&matplotlib:在Jupyter笔记本中进行三维绘图交互