Skip to content

Latest commit

 

History

History
108 lines (73 loc) · 3.15 KB

README.md

File metadata and controls

108 lines (73 loc) · 3.15 KB

groundhog

Travis Build Status

groundhog is a service and client libraries to access NASA's publicly available surface elevation data, SRTM. Given a table with GPS information, this package provides utilities to enrich that dataset with elevation and slope features. Just load the container and one of the clients, and you're all set!


Getting Started

Clone this repository.

git clone https://github.com/uptake/groundhog.git
cd groundhog

Building the container

We'll put the container up on Dockerhub soon (status). For now, you can build locally from this repo.

docker build -t groundhog -f Dockerfile .

Running the app

Run this command to kick up a local version of the app

docker run -p 5005:5005 --name groundhog_server groundhog

You can now hit the API on localhost. Run the following command in a separate terminal to see the help docs.

curl http://localhost:5005

To stop the container by name (if you used the --name tag when launching it), do the following:

docker stop groundhog_server

Clients

Given a running instance of the service, choose a client to get and append elevation data to your dataset.

R client

We provide an R client for the service. Given a data.table with GPS information and (optionally) bearing, this client will use groundhog to enrich that dataset with elevation and slope features.

You can install the package from source

R CMD INSTALL clients/r-client/

To test it out, spin up a local version of the service, then run this example:

library(data.table)

someDT <- data.table::data.table(
    longitude = runif(10, -110, -109)
    , latitude = runif(10, 45, 46)
    , dateTime = seq.POSIXt(from = as.POSIXct("2017-01-01", tz = "UTC")
                             , to = as.POSIXct("2017-01-15", tz = "UTC")
                             , length.out = 10)
    , assetId = c(rep("ABC", 5), rep("DEF", 5))
)

groundhog::append_slope_features(someDT, hostName = "localhost", port = 5005)

Python client

We provide a Python client for the service. Given a pandas DataFrame with GPS information and (optionally) bearing, this client will use groundhog to enrich that dataset with elevation and slope features.

You can install the package from source

pushd clients/py-client
    pip install .
popd

To test it out, spin up a local version of the service, then run this example:

import pandas as pd
import numpy as np
import groundhog as gh

some_df = pd.DataFrame({
    "longitude": -110 + np.random.rand(10),
    "latitude": 45 + np.random.rand(10),
    "dateTime": pd.date_range(pd.datetime.today(), periods=10),
    "assetId": ["ABC"]*5 + ["DEF"]*5
    })

gh.append_slope_features(some_df, host_name="localhost", port=5005)

Background

This project is built on top of srtm.py, a Python library that makes the SRTM data accessible and easy to query.