Skip to content

A small set of Python functions to draw pretty maps from OpenStreetMap data. Based on osmnx, matplotlib and shapely libraries.

License

Notifications You must be signed in to change notification settings

twointheboat/prettymaps

 
 

Repository files navigation

prettymaps

A Python package to draw maps with customizable styles from OpenStreetMap data. Created using the osmnx, matplotlib, shapely and vsketch packages.

This work is licensed under a GNU Affero General Public License v3.0 (you can make commercial use, distribute and modify this project, but must disclose the source code with the license and copyright notice)

Buy Me a Coffee at ko-fi.com

Installation

Install using with pip:

pip install prettymaps

Note about crediting and NFTs:

  • Please keep the printed message on the figures crediting my repository and OpenStreetMap (mandatory by their license).
  • I am personally against NFTs for their environmental impact, the fact that they're a giant money-laundering pyramid scheme and the structural incentives they create for theft in the open source and generative art communities.
  • I do not authorize in any way this project to be used for selling NFTs, although I cannot legally enforce it. Respect the creator.
  • The AeternaCivitas and geoartnft projects have used this work to sell NFTs and refused to credit it. See how they reacted after being exposed: AeternaCivitas, geoartnft.
  • I have closed my other generative art projects on Github and won't be sharing new ones as open source to protect me from the NFT community.

As seen on Hacker News:

Tutorial

Plotting with prettymaps is very simple. Run:

prettymaps.plot(your_query)

your_query can be:

  1. An address (Example: "Porto Alegre"),
  2. Latitude / Longitude coordinates (Example: (-30.0324999, -51.2303767))
  3. A custom boundary in GeoDataFrame format
import prettymaps

plot = prettymaps.plot('Stad van de Zon, Heerhugowaard, Netherlands')

png

You can also choose from different "presets" (parameter combinations saved in JSON files)

See below an example using the "minimal" preset

plot = prettymaps.plot(
    'Stad van de Zon, Heerhugowaard, Netherlands',
    preset = 'minimal'
)

png

Run

prettymaps.presets()

to list all available presets:

preset params
0 barcelona {'layers': {'perimeter': {'circle': False}, 's...
1 barcelona-plotter {'layers': {'streets': {'width': {'primary': 5...
2 cb-bf-f {'layers': {'streets': {'width': {'trunk': 6, ...
3 default {'layers': {'perimeter': {}, 'streets': {'widt...
4 heerhugowaard {'layers': {'perimeter': {}, 'streets': {'widt...
5 macao {'layers': {'perimeter': {}, 'streets': {'cust...
6 minimal {'layers': {'perimeter': {}, 'streets': {'widt...
7 tijuca {'layers': {'perimeter': {}, 'streets': {'widt...

To examine a specific preset, run:

prettymaps.preset('default')
layers style circle radius
perimeter {}
fill: false
lw: 0
zorder: 0
null
...
500
...
streets width:
cycleway: 3.5
footway: 1
motorway: 5
pedestrian: 2
primary: 4.5
residential: 3
secondary: 4
service: 2
tertiary: 3.5
trunk: 5
unclassified: 2
alpha: 1
ec: '#475657'
fc: '#2F3737'
lw: 0
zorder: 4
building tags:
building: true
landuse: construction
ec: '#2F3737'
lw: 0.5
palette:
- '#433633'
- '#FF5E5B'
zorder: 5
water tags:
natural:
- water
- bay
ec: '#2F3737'
fc: '#a8e1e6'
hatch: ooo...
hatch_c: '#9bc3d4'
lw: 1
zorder: 3
forest tags:
landuse: forest
ec: '#2F3737'
fc: '#64B96A'
lw: 1
zorder: 2
green tags:
landuse:
- grass
- orchard
leisure: park
natural:
- island
- wood
ec: '#2F3737'
fc: '#8BB174'
hatch: ooo...
hatch_c: '#A7C497'
lw: 1
zorder: 1
beach tags:
natural: beach
ec: '#2F3737'
fc: '#FCE19C'
hatch: ooo...
hatch_c: '#d4d196'
lw: 1
zorder: 3
parking tags:
amenity: parking
highway: pedestrian
man_made: pier
ec: '#2F3737'
fc: '#F2F4CB'
lw: 1
zorder: 3
background .nan
...
fc: '#F2F4CB'
zorder: -1

Insted of using the default configuration you can customize several parameters. The most important are:

  • layers: A dictionary of OpenStreetMap layers to fetch.
    • Keys: layer names (arbitrary)
    • Values: dicts representing OpenStreetMap queries
  • style: Matplotlib style parameters
    • Keys: layer names (the same as before)
    • Values: dicts representing Matplotlib style parameters
plot = prettymaps.plot(
    # Your query. Example: "Porto Alegre" or (-30.0324999, -51.2303767) (GPS coords)
    your_query,
    # Dict of OpenStreetMap Layers to plot. Example:
    # {'building': {'tags': {'building': True}}, 'water': {'tags': {'natural': 'water'}}}
    # Check the /presets folder for more examples
    layers,
    # Dict of style parameters for matplotlib. Example:
    # {'building': {'palette': ['#f00','#0f0','#00f'], 'edge_color': '#333'}}
    style,
    # Preset to load. Options include:
    # ['default', 'minimal', 'macao', 'tijuca']
    preset,
    # Save current parameters to a preset file.
    # Example: "my-preset" will save to "presets/my-preset.json"
    save_preset,
    # Whether to update loaded preset with additional provided parameters. Boolean
    update_preset,
    # Plot with circular boundary. Boolean
    circle,
    # Plot area radius. Float
    radius,
    # Dilate the boundary by this amount. Float
    dilate
)

plot is a python dataclass containing:

@dataclass
class Plot:
    # A dictionary of GeoDataFrames (one for each plot layer)
    geodataframes: Dict[str, gp.GeoDataFrame]
    # A matplotlib figure
    fig: matplotlib.figure.Figure
    # A matplotlib axis object
    ax: matplotlib.axes.Axes

Here's an example of running prettymaps.plot() with customized parameters:

plot = prettymaps.plot(
    'Praça Ferreira do Amaral, Macau',
    circle = True,
    radius = 1100,
    layers = {
        "green": {
            "tags": {
                "landuse": "grass",
                "natural": ["island", "wood"],
                "leisure": "park"
            }
        },
        "forest": {
            "tags": {
                "landuse": "forest"
            }
        },
        "water": {
            "tags": {
                "natural": ["water", "bay"]
            }
        },
        "parking": {
            "tags": {
                "amenity": "parking",
                "highway": "pedestrian",
                "man_made": "pier"
            }
        },
        "streets": {
            "width": {
                "motorway": 5,
                "trunk": 5,
                "primary": 4.5,
                "secondary": 4,
                "tertiary": 3.5,
                "residential": 3,
            }
        },
        "building": {
            "tags": {"building": True},
        },
    },
    style = {
        "background": {
            "fc": "#F2F4CB",
            "ec": "#dadbc1",
            "hatch": "ooo...",
        },
        "perimeter": {
            "fc": "#F2F4CB",
            "ec": "#dadbc1",
            "lw": 0,
            "hatch": "ooo...",
        },
        "green": {
            "fc": "#D0F1BF",
            "ec": "#2F3737",
            "lw": 1,
        },
        "forest": {
            "fc": "#64B96A",
            "ec": "#2F3737",
            "lw": 1,
        },
        "water": {
            "fc": "#a1e3ff",
            "ec": "#2F3737",
            "hatch": "ooo...",
            "hatch_c": "#85c9e6",
            "lw": 1,
        },
        "parking": {
            "fc": "#F2F4CB",
            "ec": "#2F3737",
            "lw": 1,
        },
        "streets": {
            "fc": "#2F3737",
            "ec": "#475657",
            "alpha": 1,
            "lw": 0,
        },
        "building": {
            "palette": [
                "#FFC857",
                "#E9724C",
                "#C5283D"
            ],
            "ec": "#2F3737",
            "lw": 0.5,
        }
    }
)

png

In order to plot an entire region and not just a rectangular or circular area, set

radius = False
plot = prettymaps.plot(
    'Bom Fim, Porto Alegre, Brasil', radius = False,
)

png

You can access layers's GeoDataFrames directly like this:

# Run prettymaps in show = False mode (we're only interested in obtaining the GeoDataFrames)
plot = prettymaps.plot('Centro Histórico, Porto Alegre', show = False)
plot.geodataframes['building']
addr:housenumber addr:street amenity operator website geometry addr:postcode name office opening_hours ... contact:phone bus public_transport source:name government ways name:fr type building:part architect
element_type osmid
node 2407915698 820 Rua Washington Luiz NaN NaN NaN POINT (-51.23212 -30.03670) 90010-460 NaN NaN NaN ... NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
way 126665330 387 Rua dos Andradas place_of_worship NaN NaN POLYGON ((-51.23518 -30.03275, -51.23512 -30.0... 90020-002 Igreja Nossa Senhora das Dores NaN NaN ... NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
126665331 1001 Rua dos Andradas NaN NaN http://www.ruadapraiashopping.com.br POLYGON ((-51.23167 -30.03066, -51.23160 -30.0... 90020-015 Rua da Praia Shopping NaN Mo-Fr 09:00-21:00; Sa 08:00-20:00 ... NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
129176990 1020 Rua 7 de Setembro NaN NaN http://www.memorial.rs.gov.br POLYGON ((-51.23117 -30.02891, -51.23120 -30.0... 90010-191 Memorial do Rio Grande do Sul NaN Tu-Sa 10:00-18:00 ... NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
129176991 NaN Praça da Alfândega NaN NaN http://www.margs.rs.gov.br POLYGON ((-51.23153 -30.02914, -51.23156 -30.0... 90010-150 Museu de Arte do Rio Grande do Sul NaN Tu-Su 10:00-19:00 ... NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...
relation 6760281 NaN NaN NaN NaN NaN POLYGON ((-51.23238 -30.03337, -51.23223 -30.0... NaN NaN NaN NaN ... NaN NaN NaN NaN NaN [457506887, 457506886] NaN multipolygon NaN NaN
6760282 NaN NaN NaN NaN NaN POLYGON ((-51.23203 -30.03340, -51.23203 -30.0... NaN Atheneu Espírita Cruzeiro do Sul NaN NaN ... NaN NaN NaN NaN NaN [457506875, 457506889, 457506888] NaN multipolygon NaN NaN
6760283 NaN NaN NaN NaN NaN POLYGON ((-51.23284 -30.03367, -51.23288 -30.0... NaN Palacete Chaves NaN NaN ... NaN NaN NaN NaN NaN [457506897, 457506896] NaN multipolygon NaN Theodor Wiederspahn
6760284 NaN NaN NaN NaN NaN POLYGON ((-51.23499 -30.03412, -51.23498 -30.0... NaN NaN NaN NaN ... NaN NaN NaN NaN NaN [457506910, 457506913] NaN multipolygon NaN NaN
14393526 1044 Rua Siqueira Campos NaN NaN https://www.sefaz.rs.gov.br POLYGON ((-51.23125 -30.02813, -51.23128 -30.0... NaN Secretaria Estadual da Fazenda NaN NaN ... NaN NaN NaN NaN NaN [236213286, 1081974882] NaN multipolygon NaN NaN

2423 rows × 105 columns

Search a building by name and display it:

plot.geodataframes['building'][
        plot.geodataframes['building'].name == 'Catedral Metropolitana Nossa Senhora Mãe de Deus'
].geometry[0]

svg

Plot mosaic of building footprints

import numpy as np
import osmnx as ox
from matplotlib import pyplot as plt
from matplotlib.font_manager import FontProperties

# Run prettymaps in show = False mode (we're only interested in obtaining the GeoDataFrames)
plot = prettymaps.plot('Porto Alegre', show = False)
# Get list of buildings from plot's geodataframes dict
buildings = plot.geodataframes['building']
# Project from lat / long
buildings = ox.project_gdf(buildings)
buildings = [b for b in buildings.geometry if b.area > 0]

# Draw Matplotlib mosaic of n x n building footprints
n = 6
fig,axes = plt.subplots(n,n, figsize = (7,6))
# Set background color
fig.patch.set_facecolor('#5cc0eb')
# Figure title
fig.suptitle(
    'Buildings of Porto Alegre',
    size = 25,
    color = '#fff',
    fontproperties = FontProperties(fname = '../assets/PermanentMarker-Regular.ttf')
)
# Draw each building footprint on a separate axis
for ax,building in zip(np.concatenate(axes),buildings):
    ax.plot(*building.exterior.xy, c = '#ffffff')
    ax.autoscale(); ax.axis('off'); ax.axis('equal')

png

Access plot.ax or plot.fig to add new elements to the matplotlib plot:

from matplotlib.font_manager import FontProperties

plot = prettymaps.plot(
    (41.39491,2.17557),
    preset = 'barcelona',
)

# Change background color
plot.fig.patch.set_facecolor('#F2F4CB')
# Add title
plot.ax.set_title(
    'Barcelona',
    fontproperties = FontProperties(
        fname = '../assets/PermanentMarker-Regular.ttf',
        size = 50
    )
)

plt.show()

png

Use plotter mode to export a pen plotter-compatible SVG (thanks to abey79's amazing vsketch library)

plot = prettymaps.plot(
    (41.39491,2.17557),
    mode = 'plotter',
    layers = dict(perimeter = {}),
    preset = 'barcelona-plotter',
    scale_x = .6,
    scale_y = -.6,
)

png

Some other examples

plot = prettymaps.plot(
    # City name
    'Barra da Tijuca',
    dilate = 0,
    figsize = (22,10),
    preset = 'tijuca',
)

png

plot = prettymaps.plot(
    'Stad van de Zon, Heerhugowaard, Netherlands',
    preset = 'heerhugowaard',
)

png

Use prettymaps.create_preset() to create a preset:

prettymaps.create_preset(
    "my-preset",
    layers = {
        "building": {
            "tags": {
                "building": True,
                "leisure": [
                    "track",
                    "pitch"
                ]
            }
        },
        "streets": {
            "width": {
                "trunk": 6,
                "primary": 6,
                "secondary": 5,
                "tertiary": 4,
                "residential": 3.5,
                "pedestrian": 3,
                "footway": 3,
                "path": 3
            }
        },
    },
    style = {
        "perimeter": {
            "fill": False,
            "lw": 0,
            "zorder": 0
        },
        "streets": {
            "fc": "#F1E6D0",
            "ec": "#2F3737",
            "lw": 1.5,
            "zorder": 3
        },
        "building": {
            "palette": [
                "#fff"
            ],
            "ec": "#2F3737",
            "lw": 1,
            "zorder": 4
        }
    }
)

prettymaps.preset('my-preset')
layers style circle radius dilate
building tags:
building: true
leisure:
- track
- pitch
ec: '<span style="background-color:#2F3737; color:#fff">#2F3737'
lw: 1
palette:
- '#fff'
zorder: 4
null
...
null
...
null
...
streets width:
footway: 3
path: 3
pedestrian: 3
primary: 6
residential: 3.5
secondary: 5
tertiary: 4
trunk: 6
ec: '#2F3737'
fc: '#F1E6D0'
lw: 1.5
zorder: 3
perimeter .nan
...
fill: false
lw: 0
zorder: 0

Use prettymaps.delete_preset() to delete presets:

# Show presets before deletion
print('Before deletion:')
display(prettymaps.presets())
# Delete 'my-preset'
prettymaps.delete_preset('my-preset')
# Show presets after deletion
print('After deletion:')
display(prettymaps.presets())
Before deletion:
preset params
0 barcelona {'layers': {'perimeter': {'circle': False}, 's...
1 barcelona-plotter {'layers': {'streets': {'width': {'primary': 5...
2 cb-bf-f {'layers': {'streets': {'width': {'trunk': 6, ...
3 default {'layers': {'perimeter': {}, 'streets': {'widt...
4 heerhugowaard {'layers': {'perimeter': {}, 'streets': {'widt...
5 macao {'layers': {'perimeter': {}, 'streets': {'cust...
6 minimal {'layers': {'perimeter': {}, 'streets': {'widt...
7 my-preset {'layers': {'building': {'tags': {'building': ...
8 tijuca {'layers': {'perimeter': {}, 'streets': {'widt...
After deletion:
preset params
0 barcelona {'layers': {'perimeter': {'circle': False}, 's...
1 barcelona-plotter {'layers': {'streets': {'width': {'primary': 5...
2 cb-bf-f {'layers': {'streets': {'width': {'trunk': 6, ...
3 default {'layers': {'perimeter': {}, 'streets': {'widt...
4 heerhugowaard {'layers': {'perimeter': {}, 'streets': {'widt...
5 macao {'layers': {'perimeter': {}, 'streets': {'cust...
6 minimal {'layers': {'perimeter': {}, 'streets': {'widt...
7 tijuca {'layers': {'perimeter': {}, 'streets': {'widt...

Use prettymaps.multiplot and prettymaps.Subplot to draw multiple regions on the same canvas

# Draw several regions on the same canvas
prettymaps.multiplot(
    prettymaps.Subplot(
        'Cidade Baixa, Porto Alegre',
        style={'building': {'palette': ['#49392C', '#E1F2FE', '#98D2EB']}}
    ),
    prettymaps.Subplot(
        'Bom Fim, Porto Alegre',
        style={'building': {'palette': ['#BA2D0B', '#D5F2E3', '#73BA9B', '#F79D5C']}}
    ),
    prettymaps.Subplot(
        'Farroupilha, Porto Alegre',
        style={'building': {'palette': ['#EEE4E1', '#E7D8C9', '#E6BEAE']}}
    ),
    # Load a global preset
    preset='cb-bf-f',
    # Figure size
    figsize=(12, 12)
)

png

About

A small set of Python functions to draw pretty maps from OpenStreetMap data. Based on osmnx, matplotlib and shapely libraries.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Jupyter Notebook 99.6%
  • Other 0.4%