‘EddiesObservations‘ object has no attribute ‘filled‘

问题

最近用到了py-eddy-tracker, 参照官方文档进行漩涡区域填充时发现调用py_eddy_tracker.observations.observation.EddiesObservations.filled 报错:

'EddiesObservations' object has no attribute 'filled'

尝试

filled函数的官方文档
查看官网提供的源码发现是有filled函数的,但是自己本地的库中没有.

检查版本发现本地版本为3.2.0, 官方文档版本为3.5.0

> pip show pyEddyTracker

Name: pyEddyTracker
Version: 3.2.0
...

而且pip无法安装3.5.0

> pip install pyEddyTracker==3.5.0

ERROR: Could not find a version that satisfies the requirement pyEddyTracker==3.5.0 (from versions: 3.0.0, 3.1.0, 3.2.0)
ERROR: No matching distribution found for pyEddyTracker==3.5.0

尝试使用文档提供的办法手动安装最新版本:
最后提示版本过低, 我的python3.6无法安装最新版的py-eddy-tracker

RuntimeError: Cannot install on Python version 3.6.4; only versions >=3.7,<3.10 are supported

解决方法

升级python
升级python版本是不可能的, 这辈子都不可能的.

手动修改observation.py源码

# 32行+
from matplotlib.cm import get_cmap
from matplotlib.collections import PolyCollection
from matplotlib.colors import Normalize

# 1259行+
    def parse_varname(self, name):
        return self[name] if isinstance(name, str) else name

    def format_label(self, label):
        t0, t1 = self.period
        return label.format(
            t0=t0,
            t1=t1,
            nb_obs=len(self),
        )

    def filled(
            self,
            ax,
            varname=None,
            ref=None,
            intern=False,
            cmap="magma_r",
            lut=10,
            vmin=None,
            vmax=None,
            factor=1,
            **kwargs,
        ):
        """
        :param matplotlib.axes.Axes ax: matplotlib axe used to draw
        :param str,array,None varname: variable used to fill the contours, or an array of same size than obs
        :param float,None ref: if defined, all coordinates are wrapped with ref as western boundary
        :param bool intern: if True draw speed contours instead of effective contours
        :param str cmap: matplotlib colormap name
        :param int,None lut: Number of colors in the colormap
        :param float,None vmin: Min value of the colorbar
        :param float,None vmax: Max value of the colorbar
        :param float factor: multiply value by
        :return: Collection drawed
        :rtype: matplotlib.collections.PolyCollection

        .. minigallery:: py_eddy_tracker.EddiesObservations.filled
        """
        x_name, y_name = self.intern(intern)
        x, y = self[x_name], self[y_name]
        if ref is not None:
            # TODO : maybe buggy with global display
            shape_out = x.shape
            x, y = wrap_longitude(x.reshape(-1), y.reshape(-1), ref)
            x, y = x.reshape(shape_out), y.reshape(shape_out)
        verts = list()
        for x_, y_ in zip(x, y):
            verts.append(create_vertice(x_, y_))
        if "facecolors" not in kwargs:
            kwargs = kwargs.copy()
            cmap = get_cmap(cmap, lut)
            v = self.parse_varname(varname) * factor
            if vmin is None:
                vmin = v.min()
            if vmax is None:
                vmax = v.max()
            v = (v - vmin) / (vmax - vmin)
            colors = [cmap(v_) for v_ in v]
            kwargs["facecolors"] = colors
        if "label" in kwargs:
            kwargs["label"] = self.format_label(kwargs["label"])
        c = PolyCollection(verts, **kwargs)
        ax.add_collection(c)
        c.cmap = cmap
        c.norm = Normalize(vmin=vmin, vmax=vmax)
        return c

结果

成功调用

fig = plt.figure(figsize=(3, 2))
ax = fig.add_axes([.03, .03, .94, .94])
ax.set_ylim(4, 30)
ax.set_xlim(105.5, 150)
plt.axis('off')
plt.gca().xaxis.set_major_locator(plt.NullLocator())
plt.gca().yaxis.set_major_locator(plt.NullLocator())
plt.subplots_adjust(top=1, bottom=0, right=1, left=0, hspace=0, wspace=0)
plt.margins(0, 0)
ax.set_aspect('equal')
a.filled(ax, 'amplitude', vmin=0, vmax=.5, color='b')     # 填充
c.filled(ax, 'amplitude', vmin=0, vmax=.5, color='r')
plt.show()