Static maps¶

Over the course of the last weeks, we have already become familiar with plotting basic static maps using geopandas.GeoDataFrame.plot(), for instance in lessons 2, 3, and 4. We also learned that geopandas.GeoDataFrame.plot() uses the matplotlib.pyplot library, and that most of its arguments and options are accepted by geopandas.

To refresh our memory about the basics of plotting maps, let's create a static accessibility map of the Helsinki metropolitan area, that also shows roads and metro lines (three layers, overlaid onto each other). Remember that the input data sets need to be in the same coordinate system!

Data¶

We will use three different data sets:

the travel time to the Helsinki railway station we used in lesson 4, which is in DATA_DIRECTORY / "helsinki_region_travel_times_to_railway_station",
the Helsinki Metro network, available via WFS from the city's map services, and
a simplified network of the most important roads in the metropolitan region, also available via WFS from the same endpoint.

In [1]:

import pathlib
NOTEBOOK_PATH = pathlib.Path().resolve()
DATA_DIRECTORY = NOTEBOOK_PATH / "data"

In [2]:

import geopandas
import numpy
import ssl
ssl._create_default_https_context = ssl._create_unverified_context

accessibility_grid = geopandas.read_file(
    DATA_DIRECTORY
    / "helsinki_region_travel_times_to_railway_station"
    / "helsinki_region_travel_times_to_railway_station.gpkg"
)
accessibility_grid["pt_r_t"] = accessibility_grid["pt_r_t"].replace(-1, numpy.nan)

WFS_BASE_URL = (
    "https://kartta.hel.fi/ws/geoserver/avoindata/wfs"
    "?service=wfs"
    "&version=2.0.0"
    "&request=GetFeature"
    "&srsName=EPSG:3879"
    "&typeName={layer:s}"
)

metro = (
    geopandas.read_file(
        WFS_BASE_URL.format(layer="avoindata:Seutukartta_liikenne_metro_rata")
    )
    .set_crs("EPSG:3879")
)
roads = (
    geopandas.read_file(
        WFS_BASE_URL.format(layer="avoindata:Seutukartta_liikenne_paatiet")
    )
    .set_crs("EPSG:3879")
)

Attention: Coordinate Reference Systems

Remember that different geo-data frames need to be in the same coordinate system before plotting them in the same map. geopandas.GeoDataFrame.plot() does not reproject data automatically.

You can always check it with a simple assert statement.

In [3]:

assert accessibility_grid.crs == metro.crs == roads.crs, "Input data sets’ CRS differs"

---------------------------------------------------------------------------
AssertionError                            Traceback (most recent call last)
Cell In[3], line 1
----> 1 assert accessibility_grid.crs == metro.crs == roads.crs, "Input data sets’ CRS differs"

AssertionError: Input data sets’ CRS differs

If multiple data sets do not share a common CRS, first, figure out which CRS they have assigned (if any!), then transform the data into a common reference system:

In [4]:

print(accessibility_grid.crs)

EPSG:3067

In [5]:

print(metro.crs)

EPSG:3879

In [6]:

print(roads.crs)

EPSG:3879

In [7]:

roads = roads.to_crs(accessibility_grid.crs)
metro = metro.to_crs(accessibility_grid.crs)

In [8]:

assert accessibility_grid.crs == metro.crs == roads.crs, "Input data sets’ CRS differs"

Plotting a multi-layer map¶

Hint: Check Your Understanding

Complete the next steps at your own pace (clear out the code cells first). Make sure to revisit previous lessons if you feel unsure how to complete a task.

Visualize a multi-layer map using the geopandas.GeoDataFrame.plot() method;

First, plot the accessibility grid using a ‘quantiles’ classification scheme;

Then, add roads network and metro lines in the same plot (remember the ax parameter).

Remember the following options that can be passed to plot():

style the polygon layer:
- define a classification scheme using the scheme parameter
- change the colour map using cmap
- control the layer’s transparency the alpha parameter (0 is fully transparent, 1 fully opaque)
style the line layers:
- adjust the line colour using the color parameter
- change the linewidth, as needed

The layers have different extents (roads covers a much larger area). You can use the axes’ (ax) methods set_xlim() and set_ylim() to set the horizontal and vertical extents of the map (e.g., to a geo-data frame’s total_bounds).

In [9]:

import matplotlib.pyplot as plt
ax = accessibility_grid.plot(
    figsize=(12, 8),

    column="pt_r_t",
    scheme="quantiles",
    cmap="Spectral",
    linewidth=0,
    alpha=0.8
)

metro.plot(
    ax=ax,
    color="orange",
    linewidth=2.5
)

roads.plot(
    ax=ax,
    color="grey",
    linewidth=0.8
)

minx, miny, maxx, maxy = accessibility_grid.total_bounds
ax.set_xlim(minx, maxx)
ax.set_ylim(miny, maxy)

Out[9]:

(6665250.00004393, 6698000.000038021)

No description has been provided for this image

Accessibility grid plot

Adding a legend¶

To plot a legend for a map, add the legend=True parameter.

For figures without a classification scheme, the legend consists of a colour gradient bar. The legend is an instance of matplotlib.pyplot.colorbar.Colorbar, and all arguments defined in legend_kwds are passed through to it. See below how to use the label property to set the legend title:

In [10]:

ax = accessibility_grid.plot(
    figsize=(12, 8),

    column="pt_r_t",
    cmap="Spectral",
    linewidth=0,
    alpha=0.8,

    legend=True,
    legend_kwds={"label": "Travel time (min)"}
)

Accessibility grid plot

Hint: Set Other Colorbar Parameters

Check out matplotlib.pyplot.colorbar.Colorbar’s documentation and experiment with other parameters! Anything you add to the legend_kwds dictionary will be passed to the color bar.

For figures that use a classification scheme, on the other hand, plot() creates a matplotlib.legend.Legend. Again, legend_kwds are passed through, but the parameters are slightly different: for instance, use title instead of label to set the legend title:

In [11]:

accessibility_grid.plot(
    figsize=(12, 8),

    column="pt_r_t",
    scheme="quantiles",
    cmap="Spectral",
    linewidth=0,
    alpha=0.8,

    legend=True,
    legend_kwds={"title": "Travel time (min)"}
)

Out[11]:

<Axes: >

Hint: Set Other Legend Parameters

Check out matplotlib.pyplot.legend.Legend’s documentation, and experiment with other parameters! Anything you add to the legend_kwds dictionary will be passed to the legend.

What legend_kwds keyword would spread the legend onto two columns?

Adding a base map¶

For better orientation, it is often helpful to add a base map to a map plot. A base map, for instance, from map providers such as OpenStreetMap or Stamen, adds streets, place names, and other contextual information.

The Python package contextily takes care of downloading the necessary map tiles and rendering them in a geopandas plot.

Caution: Web Mercator

Map tiles from online map providers are typically in Web Mercator projection (EPSG:3857). It is generally advisable to transform all other layers to EPSG:3857, too.

In [12]:

accessibility_grid = accessibility_grid.to_crs("EPSG:3857")
metro = metro.to_crs("EPSG:3857")
roads = roads.to_crs("EPSG:3857")

To add a base map to an existing plot, use the contextily.add_basemap() function, and supply the plot’s ax object obtained in an earlier step.

In [13]:

import contextily

ax = accessibility_grid.plot(
    figsize=(12, 8),

    column="pt_r_t",
    scheme="quantiles",
    cmap="Spectral",
    linewidth=0,
    alpha=0.8,

    legend=True,
    legend_kwds={"title": "Travel time (min)"}
)
contextily.add_basemap(ax, source=contextily.providers.OpenStreetMap.Mapnik)

Static-maps

Static-maps#

Static maps¶

Data¶

Plotting a multi-layer map¶

Adding a legend¶

Adding a base map¶