degaussverse 1.0 updates (#1)

.dockerignore

@@ -3,6 +3,4 @@
 
 # except what we need
 !/renv.lock
-!/tracts_2010_sf_5072.rds
-!/tract_dep_index_2018.rds
-!/dep_index.R
+!/entrypoint.R

.github/workflows/build-deploy-pr.yaml

@@ -0,0 +1,40 @@
+name: build-deploy-pr
+on:
+  pull_request:
+jobs:
+  deploy-images:
+    runs-on: ubuntu-latest
+    env:
+      registry: ghcr.io
+      username: degauss-org
+      repository: dep_index
+    strategy:
+      fail-fast: false
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v2
+      - name: create latest tag variable
+        run: |
+          container="${{ env.registry }}/${{ env.username}}/${{ env.repository }}:latest"
+          echo "container=${container}" >> $GITHUB_ENV
+      - name: create pull request tag variable based on name of associated branch
+        if: github.event_name == 'pull_request'
+        run: |
+          versioned="${{ env.registry }}/${{ env.username}}/${{ env.repository }}:${GITHUB_HEAD_REF}"
+          echo "versioned=${versioned}" >> $GITHUB_ENV
+      - name: build container
+        run: |
+          docker build -t ${{ env.container }} .
+      - name: test run container
+        run: |
+          docker run --rm -v "${PWD}/test":/tmp ${{ env.container }} my_address_file_geocoded.csv
+      - name: login to ghcr
+        uses: docker/login-action@v1
+        with:
+          registry: ${{ env.registry }}
+          username: ${{ env.username }}
+          password: ${{ secrets.GITHUB_TOKEN }}
+      - name: deploy pull request container
+        run: |
+          docker tag ${{ env.container }} ${{ env.versioned }}
+          docker push ${{ env.versioned }}

.github/workflows/build-deploy-release.yaml

@@ -0,0 +1,42 @@
+name: build-deploy-release
+on:
+  release:
+    types: [published]
+jobs:
+  deploy-images:
+    runs-on: ubuntu-latest
+    env:
+      registry: ghcr.io
+      username: degauss-org
+      repository: dep_index
+    strategy:
+      fail-fast: false
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v2
+      - name: create latest tag variable
+        run: |
+          container="${{ env.registry }}/${{ env.username}}/${{ env.repository }}:latest"
+          echo "container=${container}" >> $GITHUB_ENV
+      - name: create release tag variable
+        if: github.event_name == 'release'
+        run: |
+          versioned="${{ env.registry }}/${{ env.username}}/${{ env.repository }}:${GITHUB_REF##*/}"
+          echo "versioned=${versioned}" >> $GITHUB_ENV
+      - name: build container
+        run: |
+          docker build -t ${{ env.container }} .
+      - name: test container
+        run: |
+          docker run --rm -v "${PWD}/test":/tmp ${{ env.container }} my_address_file_geocoded.csv
+      - name: login to ghcr
+        uses: docker/login-action@v1
+        with:
+          registry: ${{ env.registry }}
+          username: ${{ env.username }}
+          password: ${{ secrets.GITHUB_TOKEN }}
+      - name: deploy release (and latest) container
+        run: |
+          docker tag ${{ env.container }} ${{ env.versioned }}
+          docker push ${{ env.versioned }}
+          docker push ${{ env.container }}

Dockerfile

@@ -1,8 +1,8 @@
-FROM rocker/r-ver:4.0.4
+FROM rocker/r-ver:4.0.5
 
 # DeGAUSS container metadata
 ENV degauss_name="dep_index"
-ENV degauss_version="0.1"
+ENV degauss_version="0.2.0"
 ENV degauss_description="census tract-level deprivation index"
 
 # add OCI labels based on environment variables too
@@ -11,33 +11,28 @@ LABEL "org.degauss.version"="${degauss_version}"
 LABEL "org.degauss.description"="${degauss_description}"
 LABEL "org.degauss.argument"="${degauss_argument}"
 
-# ADD https://geomarker.s3.us-east-2.amazonaws.com/geometries/tracts_2010_sf_5072.rds /opt/tracts_2010_sf_5072.rds
-# ADD https://geomarker.s3.us-east-2.amazonaws.com/tract_dep_index_2018.rds /opt/tract_dep_index_18.rds
+RUN R --quiet -e "install.packages('remotes', repos = c(CRAN = 'https://packagemanager.rstudio.com/all/__linux__/focal/latest'))"
 
-COPY tracts_2010_sf_5072.rds /opt/tracts_2010_sf_5072.rds
-COPY tract_dep_index_2018.rds /opt/tract_dep_index_18.rds
-
-# install required version of renv
-RUN R --quiet -e "install.packages('remotes', repos = 'https://cran.rstudio.com')"
-# make sure version matches what is used in the project: packageVersion('renv')
-ENV RENV_VERSION 0.13.2
-RUN R --quiet -e "remotes::install_github('rstudio/renv@${RENV_VERSION}')"
+RUN R --quiet -e "remotes::install_github('rstudio/[email protected]')"
 
 WORKDIR /app
 
 RUN apt-get update \
-  && apt-get install -yqq --no-install-recommends \
-  libgdal-dev \
-  libgeos-dev \
-  libudunits2-dev \
-  libproj-dev \
-  && apt-get clean
+    && apt-get install -yqq --no-install-recommends \
+    libgdal-dev \
+    libgeos-dev \
+    libudunits2-dev \
+    libproj-dev \
+    && apt-get clean
 
 COPY renv.lock .
+
 RUN R --quiet -e "renv::restore(repos = c(CRAN = 'https://packagemanager.rstudio.com/all/__linux__/focal/latest'))"
 
-COPY dep_index.R .
+ADD https://geomarker.s3.us-east-2.amazonaws.com/geometries/tracts_2010_sf_5072.rds /opt/tracts_2010_sf_5072.rds
+ADD https://geomarker.s3.us-east-2.amazonaws.com/tract_dep_index_2018.rds /opt/tract_dep_index_18.rds
+COPY entrypoint.R .
 
 WORKDIR /tmp
 
-ENTRYPOINT ["/app/dep_index.R"]
+ENTRYPOINT ["/app/entrypoint.R"]

LICENSE.md → LICENSE

@@ -592,4 +592,4 @@ proprietary programs. If your program is a subroutine library, you may consider
 more useful to permit linking proprietary applications with the library. If this is
 what you want to do, use the GNU Lesser General Public License instead of this
 License. But first, please read
-&lt;<http://www.gnu.org/philosophy/why-not-lgpl.html>&gt;.
+&lt;<http://www.gnu.org/philosophy/why-not-lgpl.html>&gt;.

Makefile

@@ -1,37 +1,13 @@
-REGISTRY_HOST=docker.io
-USERNAME=degauss
-NAME=$(shell basename "$(CURDIR)")
-IMAGE=$(REGISTRY_HOST)/$(USERNAME)/$(NAME)
-
-.PHONY: build test shell release clean
+.PHONY: build test shell clean
 
 build:
-	docker build -t $(IMAGE) .
+	docker build -t dep_index .
 
 test:
-	docker run --rm -v "${PWD}/test":/tmp $(IMAGE) my_address_file_geocoded.csv
+	docker run --rm -v "${PWD}/test":/tmp dep_index my_address_file_geocoded.csv
 
 shell:
-	docker run --rm -it --entrypoint=/bin/bash -v "${PWD}/test":/tmp $(IMAGE)
-
-release:
-ifndef VERSION
-	$(error VERSION is not set. Usage: "make release VERSION=X.X")
-endif
-ifndef DOCKER_USERNAME
-	$(error DOCKER_USERNAME is not set)
-endif
-ifndef DOCKER_PAT
-	$(error DOCKER_PAT is not set)
-endif
-	git commit -am "Release for image version $(VERSION)" --allow-empty
-	git tag -a $(VERSION) -m "${VERSION}"
-	git push origin ${VERSION}
-	git push
-	echo "${DOCKER_PAT}" | docker login -u "${DOCKER_USERNAME}" --password-stdin
-	docker tag ${IMAGE}:latest ${IMAGE}:${VERSION}
-	docker push ${IMAGE}:${VERSION}
-	docker push ${IMAGE}:latest
+	docker run --rm -it --entrypoint=/bin/bash -v "${PWD}/test":/tmp dep_index
 
 clean:
-	docker system prune -f
+	docker system prune -f

README.md

@@ -1,25 +1,42 @@
-# dep_index <a href='https://degauss-org.github.io/DeGAUSS/'><img src='DeGAUSS_hex.png' align='right' height='138.5' /></a>
+# dep_index <a href='https://degauss.org'><img src='https://github.com/degauss-org/degauss_hex_logo/raw/main/PNG/degauss_hex.png' align='right' height='138.5' /></a>
 
-> DeGAUSS container that adds census tract deprivation index to [geocoded](https://degauss.org/geocoder) addresses
+[![](https://img.shields.io/github/v/release/degauss-org/dep_index?color=469FC2&label=version&sort=semver)](https://github.com/degauss-org/dep_index/releases)
+[![container build status](https://github.com/degauss-org/dep_index/workflows/build-deploy-release/badge.svg)](https://github.com/degauss-org/dep_index/actions/workflows/build-deploy-release.yaml)
 
-[![GitHub Latest Tag](https://img.shields.io/github/v/tag/degauss-org/dep_index)](https://github.com/degauss-org/dep_index/releases)
+## Using
 
-## DeGAUSS example call
-
-If `my_address_file_geocoded.csv` is a file in the current working directory with coordinate columns named `lat` and `lon`, then
+If `my_address_file_geocoded.csv` is a file in the current working directory with coordinate columns named `lat` and `lon`, then the [DeGAUSS command](https://degauss.org/using_degauss.html#DeGAUSS_Commands):
 
 ```sh
-docker run --rm -v $PWD:/tmp ghcr.io/degauss-org/dep_index:0.1 my_address_file_geocoded.csv
+docker run --rm -v $PWD:/tmp ghcr.io/degauss-org/dep_index:0.2.0 my_address_file_geocoded.csv
 ```
 
-will produce `my_address_file_geocoded_dep_index_v0.1.csv` with added columns named `fips_tract_id`, `fraction_assisted_income`,	`fraction_high_school_edu`,	`median_income`,	`fraction_no_health_ins`,	`fraction_poverty`,	`fraction_vacant_housing`, and `dep_index`.
+will produce `my_address_file_geocoded_dep_index_0.2.0.csv` with added columns:
+
+- **`fips_tract_id`**: 2010 census tract identifier
+
+- 2018 American Community Survey variables:
 
-## Deprivation Index details
+    + **`fraction_assisted_income`**: fraction of households receiving public assistance income or food stamps or SNAP in the past 12 months
+    + **`fraction_high_school_edu`**: fraction of population 25 and older with educational attainment of at least high school graduation (includes GED equivalency)
+    + **`median_income`**: median household income in the past 12 months in 2018 inflation-adjusted dollars
+    + **`fraction_no_health_ins`**: fraction of poulation with no health insurance coverage
+    + **`fraction_poverty`**: fraction of population with income in past 12 months below poverty level
+    + **`fraction_vacant_housing`**: fraction of houses that are vacant
+
+- **`dep_index`**: composite measure of the 6 variables above
+
+## Geomarker Methods
 
 This container overlays the input latitude and longitude coordinates with 2010 census tracts, then joins with tract-level deprivation index data derived from the 2018 American Community Survey (ACS).
 
-For more information on the deprivation index, please see the [deprivation index page](https://geomarker.io/dep_index)
+For more information on the deprivation index, please see the [deprivation index page](https://geomarker.io/dep_index/).
+
+## Geomarker Data
+
+- 2010 tract shape files are stored at: [`s3://geomarker/geometries/tracts_2010_sf_5072.rds`](https://geomarker.s3.us-east-2.amazonaws.com/geometries/tracts_2010_sf_5072.rds).
+- 2018 deprivation index data is stored at: [`s3://geomarker/tract_dep_index_2018.rds`](https://geomarker.s3.us-east-2.amazonaws.com/tract_dep_index_2018.rds) and is also available for download at [https://geomarker.io/dep_index/](https://geomarker.io/dep_index/).
 
-## DeGAUSS details
+## DeGAUSS Details
 
 For detailed documentation on DeGAUSS, including general usage and installation, please see the [DeGAUSS homepage](https://degauss.org).

entrypoint.R

@@ -0,0 +1,44 @@
+#!/usr/local/bin/Rscript
+
+dht::greeting()
+
+## load libraries without messages or warnings
+withr::with_message_sink("/dev/null", library(dplyr))
+withr::with_message_sink("/dev/null", library(tidyr))
+withr::with_message_sink("/dev/null", library(sf))
+
+doc <- "
+      Usage:
+      entrypoint.R <filename>
+      "
+
+opt <- docopt::docopt(doc)
+
+## for interactive testing
+## opt <- docopt::docopt(doc, args = 'test/my_address_file_geocoded.csv')
+## opt <- docopt::docopt(doc, args = 'my_address_file_geocoded.csv')
+
+message("reading input file...")
+d <- dht::read_lat_lon_csv(opt$filename, nest_df = T, sf = T, project_to_crs = 5072)
+
+dht::check_for_column(d$raw_data, "lat", d$raw_data$lat)
+dht::check_for_column(d$raw_data, "lon", d$raw_data$lon)
+
+## add code here to calculate geomarkers
+message("reading tract shapefile...")
+tracts10 <- readRDS('/opt/tracts_2010_sf_5072.rds')
+
+message("joining to 2010 TIGER/Line+ census tracts using EPSG:5072 projection")
+d_tract <- st_join(d$d, tracts10) %>%
+  st_drop_geometry()
+
+message("reading deprivation index data...")
+dep_index18 <- readRDS('/opt/tract_dep_index_18.rds')
+
+message("joining 2018 tract-level deprivation index")
+d_tract <- left_join(d_tract, dep_index18, by = c('fips_tract_id' = 'census_tract_fips'))
+
+## merge back on .row after unnesting .rows into .row
+dht::write_geomarker_file(d = d_tract,
+                          raw_data = d$raw_data,
+                          filename = opt$filename)