请教利用python绘制站点数据

kangsite

请问哪位知道，如何绘制站点数据效果类似附件？求python或C++算法
数据格式如下：
Stid,Longitude,Latitude,Precipitation6h
1,113.059,29.4966,11.7053
2,113.061,29.497,8.96478
3,113.064,29.4974,8.9645
4,113.066,29.4978,8.96411
5,113.068,29.4982,8.96379
6,113.062,29.4898,8.94334
7,113.064,29.4902,8.94116
8,113.066,29.4905,8.94057
9,113.068,29.4909,8.94
10,113.07,29.4913,8.93962
11,113.077,29.4932,0.001
12,113.063,29.4829,8.90352
13,113.065,29.4833,8.90423
14,113.067,29.4837,8.90521
15,113.069,29.4841,8.90614
16,113.071,29.4845,8.90729
17,113.076,29.487,0.00100177
18,113.083,29.4947,11.5452
19,113.087,29.4993,11.5392
20,113.081,29.489,8.81343
21,113.064,29.4759,8.86586
22,113.066,29.4763,8.86758
23,113.068,29.4768,8.86918
24,113.07,29.4772,8.87062
25,113.072,29.4777,8.87233

我用的代码如下

def contour(x, y, z, linewidth=2, labels=None):
    """
    Plots contours for non-evenly spaced data.
    x,y,z must be 1d arrays.
    lines = # of contour lines (default 18 )
    linewidth = line width of lines (default 2 )
    """

    assert x.shape[0] == y.shape[0] == z.shape[
        0], "arrays x,y,z must be the same size"

    # make a grid that surrounds x,y support
    xi = np.linspace(x.min(), x.max(), 100)
    yi = np.linspace(y.min(), y.max(), 100)

    xx, yy = np.meshgrid(xi, yi)


    # grid the data.
    zi = griddata((x, y),
                  z, (xi[None, :], yi[:, None]),
                  method='linear',
                  fill_value=np.nan)
    # contour the gridded data, plotting dots at the randomly spaced data points.
    # plt.figure()
    # CS = plt.contour(xi, yi, zi, linewidth=2)
    CS = plt.contourf(xi, yi, zi)


    if labels:
        plt.xlabel(labels[0])
        plt.ylabel(labels[1])
    # plot data points.
# plt.scatter(x, y, c=z, s=60, alpha=0.7, edgecolors="none")
    plt.xlim(x.min(), x.max())
    plt.ylim(y.min(), y.max())
    plt.show()



绘制效果很难看

火枪裸鞋

plt.scatter()

kangsite

火枪裸鞋 发表于 2019-1-1 16:52
plt.scatter()

scatter 只能绘制点吧？这个是绘制类似contour的

slddbs

好像scipy里有插值的函数

kangsite

slddbs 发表于 2019-1-1 21:27
好像scipy里有插值的函数

不行。不知道边界如何处理，