ssb-sgis 0.1.2

GIS utility functions used at Statistics Norway.

ssb-sgis

GIS Python tools used in Statistics Norway.

sgis builds on the geopandas package and provides functions that make it easier to do advanced GIS in python. Features include network analysis, functions for exploring multiple GeoDataFrames in a layered interactive map, and vector operations like finding k-nearest neighbours, splitting lines by points, snapping and closing holes in polygons by size.

Network analysis examples

Preparing for network analysis:

import sgis as sg


roads = sg.read_parquet_url(
    "https://media.githubusercontent.com/media/statisticsnorway/ssb-sgis/main/tests/testdata/roads_oslo_2022.parquet"
)

nw = (
    sg.DirectedNetwork(roads)
    .remove_isolated()
    .make_directed_network(
        direction_col="oneway",
        direction_vals_bft=("B", "FT", "TF"),
        minute_cols=("drivetime_fw", "drivetime_bw"),
    )
)

rules = sg.NetworkAnalysisRules(weight="minutes")

nwa = sg.NetworkAnalysis(network=nw, rules=rules)

nwa

    NetworkAnalysis(
        network=DirectedNetwork(6364 km, percent_bidirectional=87),
        rules=NetworkAnalysisRules(weight=minutes, search_tolerance=250, search_factor=0, split_lines=False, ...),
        log=True, detailed_log=True,
    )

Get number of times each line segment was visited.

freq = nwa.get_route_frequencies(points.sample(75), points.sample(75))

sg.qtm(
    sg.buff(freq, 15),
    "frequency",
    scheme="naturalbreaks",
    cmap="plasma",
    title="Number of times each road was used.",
)

Fast many-to-many travel times/distances.

od = nwa.od_cost_matrix(points, points)

print(od)

            origin  destination    minutes
    0            0            0   0.000000
    1            0            1  13.039830
    2            0            2  10.902453
    3            0            3   8.297021
    4            0            4  14.742294
    ...        ...          ...        ...
    999995     999          995  11.038673
    999996     999          996  17.820664
    999997     999          997  10.288465
    999998     999          998  14.798257
    999999     999          999   0.000000

    [1000000 rows x 3 columns]

Get the area that can be reached within one or more breaks.

sa = nwa.service_area(
    points.iloc[[0]],
    breaks=np.arange(1, 11),
)

sg.qtm(sa, "minutes", k=10, title="Roads that can be reached within 1 to 10 minutes")

Get one or more route per origin-destination pair.

routes = nwa.get_k_routes(
    points.iloc[[0]], points.iloc[[1]], k=5, drop_middle_percent=50
)

sg.qtm(sg.buff(routes, 15), "k", title="Five fastest routes from A to B", legend=False)

More network analysis examples can be found here: https://github.com/statisticsnorway/ssb-sgis/blob/main/docs/network_analysis_demo_template.md

Road data for Norway can be downloaded here: https://kartkatalog.geonorge.no/metadata/nvdb-ruteplan-nettverksdatasett/8d0f9066-34f9-4423-be12-8e8523089313

Developer information

Git LFS

The data in the testdata directory is stored with Git LFS. Make sure git-lfs is installed and that you have run the command git lfs install at least once. You only need to run this once per user account.

Dependencies

Poetry is used for dependency management. Install poetry and run the command below from the root directory to install the dependencies.

poetry install --no-root

Tests

Use the following command from the root directory to run the tests:

poetry run pytest  # from root directory

Jupyter Notebooks

The files ending with _ipynb.py in the tests directory are jupyter notebooks stored as plain python files, using jupytext. To open them as Jupyter notebooks, right-click on them in JupyterLab and select Open With → Notebook.

When testing locally, start JupyterLab with this command:

poetry run jupter lab

For VS Code there are extensions for opening a python script as Jupyter Notebook, for example: Jupytext for Notebooks.

Formatting

Format the code with black and isort by running the following command from the root directory:

poetry run black .
poetry run isort .

Pre-commit hooks

We are using pre-commit hooks to make sure the code is correctly formatted and consistent before committing. Use the following command from the root directory in the repo to install the pre-commit hooks:

poetry run pre-commit install

It then checks the changed files before committing. You can run the pre-commit checks on all files by using this command:

poetry run pre-commit run --all-files

Documentation

To generate the API-documentation locally, run the following command from the root directory:

poetry run sphinx-build docs docs/_build

Then open the file docs/_build/index.html.

To check and run the docstrings examples, run this command:

poetry run xdoctest --command=all ./src/sgis