使用 GeoPandas 在地图上绘制点组会生成空白图像

Question

我最近开始使用 GeoPandas 来制作地图，发现它非常有用。我已经使用 Pandas 一段时间了，我发现迁移到 GeoPandas 相对轻松。但是，在使用 .dissolve() 函数对点进行分组后，我在地图上绘制点时遇到问题。

\n\n

基本上，我从国家统计局邮政编码目录 (ONSPD) 下载了一系列英国邮政编码数据以及相关的经度和纬度值。经度和纬度值基于 CRS WGS84。我可以转换为 CRS OSGB36 并毫无问题地在地图上绘制所有点。但是，如果我使用 .dissolve() 方法根据其他变量（例如“group1”和“groups”）对点进行分组，我将无法再绘制这些点。

\n\n

这是我迄今为止将所有点绘制在一起的代码：

\n\n

import pandas as pd\nimport geopandas as gif\nimport matplotlib.pyplot as plot\nimport shapely\n\n# Define a Pandas dataframe containing postcodes and \npostcodeDF = pd.DataFrame({\'pcd\': [\'RM175AG\', \'NP181PH\', \'LS8 1EN\', \'HG1 1XQ\', \'G11 6YB\', \'TN218AB\', \'GU138AL\', \'CV344BD\', \'YO126PH\', \'SO172WT\', \'PR2 8HN\', \'TF1 2HD\', \'M31 4FR\', \'CH460UB\', \'EX111LN\', \'TS214DX\', \'BN4 2LS\', \'FY8 1XL\', \'KA256BP\', \'DA1 1QR\'],\n                           \'ctry\': [\'E92000001\', \'W92000004\', \'E92000001\', \'E92000001\', \'S92000003\', \'E92000001\', \'E92000001\', \'E92000001\', \'E92000001\', \'E92000001\', \'E92000001\', \'E92000001\', \'E92000001\', \'E92000001\', \'E92000001\', \'E92000001\', \'E92000001\', \'E92000001\', \'S92000003\', \'E92000001\'],\n                           \'long\': [0.320423, -2.968257, -1.51314, -1.522386, -4.309171, 0.255878, -0.8502959999999999, -1.588886, -0.41805299999999995, -1.382926, -2.699804, -2.531778, -2.425061, -3.110276, -3.317868, -1.429442, -0.22178699999999998, -3.019257, -4.686457, 0.224912],\n                           \'lat\': [51.491329, 51.628333000000005, 53.840114, 54.002427000000004, 55.870383, 50.966097999999995, 51.275081, 52.280791, 54.296222, 50.921857, 53.780733999999995, 52.692078, 53.415147, 53.394830000000006, 50.758146, 54.680499, 50.84042, 53.753078, 55.749168999999995, 51.441911],\n                           \'group1\':[\'A\']*10 + [\'B\']*10,\n                           \'group2\':[True,False]*10})\n\n# Set up geodataframe, initially with CRS = WGS84 (since that matches the long and lat co-ordinates)\ncrs = {\'init\':\'epsg:4326\'}\ngeometry = [shapely.geometry.Point(xy) for xy in zip(postcodeDF[\'long\'], postcodeDF[\'lat\'])]\n\npostcodeGDF = gpd.GeoDataFrame(postcodeDF,\n                               crs = crs,\n                               geometry = geometry)\n\n# Convert geometry to OSGB36\npostcodeGDF = postcodeGDF.to_crs(epsg = 27700)\n\nprint(postcodeGDF)\n

\n\n

地理数据框包含以下信息：

\n\n

ctry group1  group2        lat      long      pcd  \\\n0   E92000001      A    True  51.491329  0.320423  RM175AG   \n1   W92000004      A   False  51.628333 -2.968257  NP181PH   \n2   E92000001      A    True  53.840114 -1.513140  LS8 1EN   \n3   E92000001      A   False  54.002427 -1.522386  HG1 1XQ   \n4   S92000003      A    True  55.870383 -4.309171  G11 6YB   \n5   E92000001      A   False  50.966098  0.255878  TN218AB   \n6   E92000001      A    True  51.275081 -0.850296  GU138AL   \n7   E92000001      A   False  52.280791 -1.588886  CV344BD   \n8   E92000001      A    True  54.296222 -0.418053  YO126PH   \n9   E92000001      A   False  50.921857 -1.382926  SO172WT   \n10  E92000001      B    True  53.780734 -2.699804  PR2 8HN   \n11  E92000001      B   False  52.692078 -2.531778  TF1 2HD   \n12  E92000001      B    True  53.415147 -2.425061  M31 4FR   \n13  E92000001      B   False  53.394830 -3.110276  CH460UB   \n14  E92000001      B    True  50.758146 -3.317868  EX111LN   \n15  E92000001      B   False  54.680499 -1.429442  TS214DX   \n16  E92000001      B    True  50.840420 -0.221787  BN4 2LS   \n17  E92000001      B   False  53.753078 -3.019257  FY8 1XL   \n18  S92000003      B    True  55.749169 -4.686457  KA256BP   \n19  E92000001      B   False  51.441911  0.224912  DA1 1QR   \n\n                                       geometry  \n0   POINT (561188.9840165515 179484.0452796911)  \n1   POINT (333075.0000681121 192612.9537310874)  \n2    POINT (432134.031689987 438316.9950631865)  \n3    POINT (431404.026064762 456371.9915770486)  \n4   POINT (255609.0244790429 666546.0027781442)  \n5   POINT (558502.0104336547 120942.0242662177)  \n6   POINT (480294.0287511199 153509.0281225364)  \n7   POINT (428143.9880141384 264818.0084207859)  \n8   POINT (503054.9928648224 490110.0245871434)  \n9    POINT (443470.0222579727 113781.050039533)  \n10  POINT (353983.9862037547 431827.9554603411)  \n11   POINT (364154.9903684837 310620.967712217)  \n12  POINT (371845.0195746602 391010.9943895991)  \n13   POINT (326267.022012488 389240.9700794323)  \n14   POINT (307141.054758316 96222.92213930591)  \n15  POINT (436886.0245473493 531865.0198253235)  \n16  POINT (525299.9996502266 106049.9600256799)  \n17  POINT (332890.0335889475 429005.9677596154)  \n18  POINT (231483.9972917534 653913.0284422053)  \n19  POINT (554726.0039838878 173783.0302315076)  \n

\n\n

可用于绘制地图：

\n\n

# Plot map\nfig, ax = plt.subplots(1,\n                       figsize = (4,5),\n                       dpi = 72,\n                       facecolor = \'lightblue\')\n\nax.set_position([0,0,1,1])   # Puts axis to edge of figure\nax.set_axis_off()            # Turns axis off so facecolour applies to axis area as well as bit around the outside\nax.get_xaxis().set_visible(False)   # Turns the x axis off so that \'invisible\' axis labels don\'t take up space\nax.get_yaxis().set_visible(False)\n\nlims = plt.axis(\'equal\')\n\n# N.B. Code to plot shapefile has been deleted for clarity\n\npostcodeGDF.plot(ax=ax)\n\nplt.show()\n

\n\n

地图（包括 shapefile 轮廓）如下所示：

\n\n

\n\n

但是，我想根据地理数据框中的其他变量对邮政编码进行分组（在本例中为变量“group1”和“group2”）（并最终为每个组绘制不同的颜色和标记\xe2\x80\x93虽然我还没有走到那一步）。我使用 .dissolve() 方法对点进行了分组。

\n\n

postcodesGroupby = postcodeGDF.dissolve(by = [\'group1\',\'group2\'])\nprint(postcodesGroupby)\n

\n\n

groupby 数据框如下所示：

\n\n

                                                        geometry       ctry\ngroup1 group2                                                                 \nA      False   (POINT (333075.0000681121 192612.9537310874), ...  W92000004   \n       True    (POINT (255609.0244790429 666546.0027781442), ...  E92000001   \nB      False   (POINT (326267.022012488 389240.9700794323), P...  E92000001   \n       True    (POINT (231483.9972917534 653913.0284422053), ...  E92000001   \n\n                     lat      long      pcd  \ngroup1 group2                                \nA      False   51.628333 -2.968257  NP181PH  \n       True    51.491329  0.320423  RM175AG  \nB      False   52.692078 -2.531778  TF1 2HD  \n       True    53.780734 -2.699804  PR2 8HN  \n

\n\n

但是，当我尝试使用以下方法绘制点时：

\n\n

postcodesGroupby.plot(ax=ax)\n

\n\n

...地图上没有出现任何点。

\n\n

我怀疑我错过了一些明显的东西，但我已经盯着代码看了一会儿，再也看不到树木了。任何有关如何解决此问题的建议将不胜感激。

\n

Answer 1

问题是 geopandas 目前还不支持绘制多点（以及dissolve将点分组为多点的方法）。事实上，您得到的是空白图像而不是良好的错误消息，这有点不幸。

但是，刚刚合并了一个 PR，以添加对绘制多点的支持： https: //github.com/geopandas/geopandas/pull/683。所以这将在下一个 geopandas 版本中起作用。

目前的解决方法是绘制各个点，但必须使用适当的分组颜色，以添加反映这些组的列：

# add a new column with an integer indicating the group number
postcodeGDF['group'] = postcodeGDF.groupby(['group1','group2']).ngroup()
postcodeGDF.plot(column='group', categorical=True, legend=True)

给出：