Pandas 0.19：执行箱线图时属性错误"unknown property color_cycle"仍然存在
Pandas 0.19: Attribute error "unknown property color_cycle" still exists while performing boxplot
与所说的 不同，AttributeError: Unknown property color_cycle 在最新的 Pandas 版本 (0.19.0-1) 中仍然存在。
就我而言，我有一个与此类似的数据框，但要长得多（3,000,000 行）：
A B C
1 05010001 17 1
2 05020001 5 1
3 05020002 11 1
4 05020003 2 1
5 05030001 86 1
6 07030001 84 2
7 07030002 10 2
8 08010001 16 3
出于某种原因，如果我实现这个例子，就没有错误。在我的实际情况下，执行简单的
df.boxplot(by='C')
引发了这场混乱：
AttributeError Traceback (most recent call last)
<ipython-input-51-5c645348f82f> in <module>()
----> 1 df.boxplot(by='C')
C:\Users\B4058846\AppData\Local\Enthought\Canopy\User\lib\site-packages\pandas\core\frame.pyc in boxplot(self, column, by, ax, fontsize, rot, grid, figsize, layout, return_type, **kwds)
5514
5515 columns = list(data.dtype.names)
-> 5516 arrays = [data[k] for k in columns]
5517 return arrays, columns
5518 else:
C:\Users\B4058846\AppData\Local\Enthought\Canopy\User\lib\site-packages\pandas\tools\plotting.pyc in boxplot(data, column, by, ax, fontsize, rot, grid, figsize, layout, return_type, **kwds)
2687 ax : Matplotlib axes object, optional
2688 fontsize : int or string
-> 2689 rot : label rotation angle
2690 figsize : A tuple (width, height) in inches
2691 grid : Setting this to True will show the grid
C:\Users\B4058846\AppData\Local\Enthought\Canopy\User\lib\site-packages\pandas\tools\plotting.pyc in _grouped_plot_by_column(plotf, data, columns, by, numeric_only, grid, figsize, ax, layout, return_type, **kwargs)
3091 >>> df = pandas.DataFrame(data, columns=list('ABCD'), index=index)
3092 >>>
-> 3093 >>> grouped = df.groupby(level='lvl1')
3094 >>> boxplot_frame_groupby(grouped)
3095 >>>
C:\Users\B4058846\AppData\Local\Enthought\Canopy\User\lib\site-packages\pandas\tools\plotting.pyc in plot_group(keys, values, ax)
2659 create a figure with the default figsize, causing the figsize=parameter to
2660 be ignored.
-> 2661 """
2662 if ax is None and len(plt.get_fignums()) > 0:
2663 ax = _gca()
C:\Users\B4058846\AppData\Local\Enthought\Canopy\User\lib\site-packages\matplotlib\__init__.pyc in inner(ax, *args, **kwargs)
1810 warnings.warn(msg % (label_namer, func.__name__),
1811 RuntimeWarning, stacklevel=2)
-> 1812 return func(ax, *args, **kwargs)
1813 pre_doc = inner.__doc__
1814 if pre_doc is None:
C:\Users\B4058846\AppData\Local\Enthought\Canopy\User\lib\site-packages\matplotlib\axes\_axes.pyc in boxplot(self, x, notch, sym, vert, whis, positions, widths, patch_artist, bootstrap, usermedians, conf_intervals, meanline, showmeans, showcaps, showbox, showfliers, boxprops, labels, flierprops, medianprops, meanprops, capprops, whiskerprops, manage_xticks)
3321 meanline=meanline, showfliers=showfliers,
3322 capprops=capprops, whiskerprops=whiskerprops,
-> 3323 manage_xticks=manage_xticks)
3324 return artists
3325
C:\Users\B4058846\AppData\Local\Enthought\Canopy\User\lib\site-packages\matplotlib\axes\_axes.pyc in bxp(self, bxpstats, positions, widths, vert, patch_artist, shownotches, showmeans, showcaps, showbox, showfliers, boxprops, whiskerprops, flierprops, medianprops, capprops, meanprops, meanline, manage_xticks)
3650 boxes.extend(dopatch(box_x, box_y, **final_boxprops))
3651 else:
-> 3652 boxes.extend(doplot(box_x, box_y, **final_boxprops))
3653
3654 # draw the whiskers
C:\Users\B4058846\AppData\Local\Enthought\Canopy\User\lib\site-packages\matplotlib\axes\_axes.pyc in doplot(*args, **kwargs)
3564 if vert:
3565 def doplot(*args, **kwargs):
-> 3566 return self.plot(*args, **kwargs)
3567
3568 def dopatch(xs, ys, **kwargs):
C:\Users\B4058846\AppData\Local\Enthought\Canopy\User\lib\site-packages\matplotlib\__init__.pyc in inner(ax, *args, **kwargs)
1810 warnings.warn(msg % (label_namer, func.__name__),
1811 RuntimeWarning, stacklevel=2)
-> 1812 return func(ax, *args, **kwargs)
1813 pre_doc = inner.__doc__
1814 if pre_doc is None:
C:\Users\B4058846\AppData\Local\Enthought\Canopy\User\lib\site-packages\matplotlib\axes\_axes.pyc in plot(self, *args, **kwargs)
1422 kwargs['color'] = c
1423
-> 1424 for line in self._get_lines(*args, **kwargs):
1425 self.add_line(line)
1426 lines.append(line)
C:\Users\B4058846\AppData\Local\Enthought\Canopy\User\lib\site-packages\matplotlib\axes\_base.pyc in _grab_next_args(self, *args, **kwargs)
384 return
385 if len(remaining) <= 3:
--> 386 for seg in self._plot_args(remaining, kwargs):
387 yield seg
388 return
C:\Users\B4058846\AppData\Local\Enthought\Canopy\User\lib\site-packages\matplotlib\axes\_base.pyc in _plot_args(self, tup, kwargs)
372 ncx, ncy = x.shape[1], y.shape[1]
373 for j in xrange(max(ncx, ncy)):
--> 374 seg = func(x[:, j % ncx], y[:, j % ncy], kw, kwargs)
375 ret.append(seg)
376 return ret
C:\Users\B4058846\AppData\Local\Enthought\Canopy\User\lib\site-packages\matplotlib\axes\_base.pyc in _makeline(self, x, y, kw, kwargs)
278 default_dict = self._getdefaults(None, kw, kwargs)
279 self._setdefaults(default_dict, kw, kwargs)
--> 280 seg = mlines.Line2D(x, y, **kw)
281 self.set_lineprops(seg, **kwargs)
282 return seg
C:\Users\B4058846\AppData\Local\Enthought\Canopy\User\lib\site-packages\matplotlib\lines.pyc in __init__(self, xdata, ydata, linewidth, linestyle, color, marker, markersize, markeredgewidth, markeredgecolor, markerfacecolor, markerfacecoloralt, fillstyle, antialiased, dash_capstyle, solid_capstyle, dash_joinstyle, solid_joinstyle, pickradius, drawstyle, markevery, **kwargs)
365 # update kwargs before updating data to give the caller a
366 # chance to init axes (and hence unit support)
--> 367 self.update(kwargs)
368 self.pickradius = pickradius
369 self.ind_offset = 0
C:\Users\B4058846\AppData\Local\Enthought\Canopy\User\lib\site-packages\matplotlib\artist.pyc in update(self, props)
854 func = getattr(self, 'set_' + k, None)
855 if func is None or not six.callable(func):
--> 856 raise AttributeError('Unknown property %s' % k)
857 func(v)
858 changed = True
AttributeError: Unknown property color_cycle
我留下了一个空白的 4 窗格图，而应该只有一个有 5 个箱线图列：
如何解决？
我已经证实需要 pandas 0.19.0-1 和 matplotlib 1.5.1-8 一起 才不会遇到此错误。
与所说的 AttributeError: Unknown property color_cycle 在最新的 Pandas 版本 (0.19.0-1) 中仍然存在。
就我而言，我有一个与此类似的数据框，但要长得多（3,000,000 行）：
A B C
1 05010001 17 1
2 05020001 5 1
3 05020002 11 1
4 05020003 2 1
5 05030001 86 1
6 07030001 84 2
7 07030002 10 2
8 08010001 16 3
出于某种原因，如果我实现这个例子，就没有错误。在我的实际情况下，执行简单的
df.boxplot(by='C')
引发了这场混乱：
AttributeError Traceback (most recent call last)
<ipython-input-51-5c645348f82f> in <module>()
----> 1 df.boxplot(by='C')
C:\Users\B4058846\AppData\Local\Enthought\Canopy\User\lib\site-packages\pandas\core\frame.pyc in boxplot(self, column, by, ax, fontsize, rot, grid, figsize, layout, return_type, **kwds)
5514
5515 columns = list(data.dtype.names)
-> 5516 arrays = [data[k] for k in columns]
5517 return arrays, columns
5518 else:
C:\Users\B4058846\AppData\Local\Enthought\Canopy\User\lib\site-packages\pandas\tools\plotting.pyc in boxplot(data, column, by, ax, fontsize, rot, grid, figsize, layout, return_type, **kwds)
2687 ax : Matplotlib axes object, optional
2688 fontsize : int or string
-> 2689 rot : label rotation angle
2690 figsize : A tuple (width, height) in inches
2691 grid : Setting this to True will show the grid
C:\Users\B4058846\AppData\Local\Enthought\Canopy\User\lib\site-packages\pandas\tools\plotting.pyc in _grouped_plot_by_column(plotf, data, columns, by, numeric_only, grid, figsize, ax, layout, return_type, **kwargs)
3091 >>> df = pandas.DataFrame(data, columns=list('ABCD'), index=index)
3092 >>>
-> 3093 >>> grouped = df.groupby(level='lvl1')
3094 >>> boxplot_frame_groupby(grouped)
3095 >>>
C:\Users\B4058846\AppData\Local\Enthought\Canopy\User\lib\site-packages\pandas\tools\plotting.pyc in plot_group(keys, values, ax)
2659 create a figure with the default figsize, causing the figsize=parameter to
2660 be ignored.
-> 2661 """
2662 if ax is None and len(plt.get_fignums()) > 0:
2663 ax = _gca()
C:\Users\B4058846\AppData\Local\Enthought\Canopy\User\lib\site-packages\matplotlib\__init__.pyc in inner(ax, *args, **kwargs)
1810 warnings.warn(msg % (label_namer, func.__name__),
1811 RuntimeWarning, stacklevel=2)
-> 1812 return func(ax, *args, **kwargs)
1813 pre_doc = inner.__doc__
1814 if pre_doc is None:
C:\Users\B4058846\AppData\Local\Enthought\Canopy\User\lib\site-packages\matplotlib\axes\_axes.pyc in boxplot(self, x, notch, sym, vert, whis, positions, widths, patch_artist, bootstrap, usermedians, conf_intervals, meanline, showmeans, showcaps, showbox, showfliers, boxprops, labels, flierprops, medianprops, meanprops, capprops, whiskerprops, manage_xticks)
3321 meanline=meanline, showfliers=showfliers,
3322 capprops=capprops, whiskerprops=whiskerprops,
-> 3323 manage_xticks=manage_xticks)
3324 return artists
3325
C:\Users\B4058846\AppData\Local\Enthought\Canopy\User\lib\site-packages\matplotlib\axes\_axes.pyc in bxp(self, bxpstats, positions, widths, vert, patch_artist, shownotches, showmeans, showcaps, showbox, showfliers, boxprops, whiskerprops, flierprops, medianprops, capprops, meanprops, meanline, manage_xticks)
3650 boxes.extend(dopatch(box_x, box_y, **final_boxprops))
3651 else:
-> 3652 boxes.extend(doplot(box_x, box_y, **final_boxprops))
3653
3654 # draw the whiskers
C:\Users\B4058846\AppData\Local\Enthought\Canopy\User\lib\site-packages\matplotlib\axes\_axes.pyc in doplot(*args, **kwargs)
3564 if vert:
3565 def doplot(*args, **kwargs):
-> 3566 return self.plot(*args, **kwargs)
3567
3568 def dopatch(xs, ys, **kwargs):
C:\Users\B4058846\AppData\Local\Enthought\Canopy\User\lib\site-packages\matplotlib\__init__.pyc in inner(ax, *args, **kwargs)
1810 warnings.warn(msg % (label_namer, func.__name__),
1811 RuntimeWarning, stacklevel=2)
-> 1812 return func(ax, *args, **kwargs)
1813 pre_doc = inner.__doc__
1814 if pre_doc is None:
C:\Users\B4058846\AppData\Local\Enthought\Canopy\User\lib\site-packages\matplotlib\axes\_axes.pyc in plot(self, *args, **kwargs)
1422 kwargs['color'] = c
1423
-> 1424 for line in self._get_lines(*args, **kwargs):
1425 self.add_line(line)
1426 lines.append(line)
C:\Users\B4058846\AppData\Local\Enthought\Canopy\User\lib\site-packages\matplotlib\axes\_base.pyc in _grab_next_args(self, *args, **kwargs)
384 return
385 if len(remaining) <= 3:
--> 386 for seg in self._plot_args(remaining, kwargs):
387 yield seg
388 return
C:\Users\B4058846\AppData\Local\Enthought\Canopy\User\lib\site-packages\matplotlib\axes\_base.pyc in _plot_args(self, tup, kwargs)
372 ncx, ncy = x.shape[1], y.shape[1]
373 for j in xrange(max(ncx, ncy)):
--> 374 seg = func(x[:, j % ncx], y[:, j % ncy], kw, kwargs)
375 ret.append(seg)
376 return ret
C:\Users\B4058846\AppData\Local\Enthought\Canopy\User\lib\site-packages\matplotlib\axes\_base.pyc in _makeline(self, x, y, kw, kwargs)
278 default_dict = self._getdefaults(None, kw, kwargs)
279 self._setdefaults(default_dict, kw, kwargs)
--> 280 seg = mlines.Line2D(x, y, **kw)
281 self.set_lineprops(seg, **kwargs)
282 return seg
C:\Users\B4058846\AppData\Local\Enthought\Canopy\User\lib\site-packages\matplotlib\lines.pyc in __init__(self, xdata, ydata, linewidth, linestyle, color, marker, markersize, markeredgewidth, markeredgecolor, markerfacecolor, markerfacecoloralt, fillstyle, antialiased, dash_capstyle, solid_capstyle, dash_joinstyle, solid_joinstyle, pickradius, drawstyle, markevery, **kwargs)
365 # update kwargs before updating data to give the caller a
366 # chance to init axes (and hence unit support)
--> 367 self.update(kwargs)
368 self.pickradius = pickradius
369 self.ind_offset = 0
C:\Users\B4058846\AppData\Local\Enthought\Canopy\User\lib\site-packages\matplotlib\artist.pyc in update(self, props)
854 func = getattr(self, 'set_' + k, None)
855 if func is None or not six.callable(func):
--> 856 raise AttributeError('Unknown property %s' % k)
857 func(v)
858 changed = True
AttributeError: Unknown property color_cycle
我留下了一个空白的 4 窗格图，而应该只有一个有 5 个箱线图列：
如何解决？
我已经证实需要 pandas 0.19.0-1 和 matplotlib 1.5.1-8 一起 才不会遇到此错误。