From 808d1045d7d47dff62cedbc88efe2ccfadfd5bb6 Mon Sep 17 00:00:00 2001
From: "Michael T. Kelbaugh" <Michael.Kelbaugh@jhuapl.edu>
Date: Mon, 2 Mar 2020 12:41:26 -0500
Subject: [PATCH] Dockerize the graph construction tool

Former-commit-id: 23f78dba4890c89fa77f93e57584f20b0932b335
---
 Makefile           |  3 +++
 graphs/Dockerfile  | 12 ++++++++++++
 graphs/build.py    | 20 ++++++++++++--------
 graphs/config.json | 19 +++++++++++++++++++
 4 files changed, 46 insertions(+), 8 deletions(-)
 create mode 100644 graphs/Dockerfile
 create mode 100644 graphs/config.json

diff --git a/Makefile b/Makefile
index 45bbfdf..8b19e17 100644
--- a/Makefile
+++ b/Makefile
@@ -18,3 +18,6 @@ clean:
 
 purge:
 	cd build && docker-compose -p simon down --rmi all
+
+graph:
+	cd graphs && docker build -t simon-graph:latest . && docker run -v `pwd`:/opt -d simon-graph:latest
diff --git a/graphs/Dockerfile b/graphs/Dockerfile
new file mode 100644
index 0000000..553ef47
--- /dev/null
+++ b/graphs/Dockerfile
@@ -0,0 +1,12 @@
+FROM ubuntu:18.04
+RUN apt-get update && apt-get install -y python3-dev=3.6.7-1~18.04 python3-pip=9.0.1-2.3~ubuntu1.18.04.1 cmake=3.10.2-1ubuntu2.18.04.1 wget && apt-get -qy autoremove && apt-get clean && rm -r /var/lib/apt/lists/*
+RUN wget https://github.com/libspatialindex/libspatialindex/releases/download/1.9.3/spatialindex-src-1.9.3.tar.gz -P /
+RUN tar xfvz /spatialindex-src-1.9.3.tar.gz
+WORKDIR /spatialindex-src-1.9.3
+RUN cmake -DCMAKE_INSTALL_PREFIX=/libspatialindex .
+RUN make
+RUN make install
+ENV LD_LIBRARY_PATH=/libspatialindex/lib
+COPY ./requirements.txt /
+RUN pip3 install -r /requirements.txt
+ENTRYPOINT ["python3", "/opt/build.py"]
diff --git a/graphs/build.py b/graphs/build.py
index 1610049..bb455c9 100644
--- a/graphs/build.py
+++ b/graphs/build.py
@@ -16,6 +16,10 @@ import itertools
 from collections import defaultdict
 import uuid
 import os
+import json
+
+with open(os.path.join(os.path.dirname(__file__), "config.json")) as f:
+    config = json.load(f)
 
 # where to save the graphs
 save_dir = os.path.join(os.path.dirname(__file__), "out")
@@ -24,19 +28,19 @@ save_dir = os.path.join(os.path.dirname(__file__), "out")
 shapefile_dir = os.path.join(os.path.dirname(__file__), "shapefiles")
 
 # the coordinate reference system the shapefiles are defined on
-projection = 3085
+projection = config["projection"]
 
 # scale the areas calculated for shapefile geometries from square kilometers to square meters
-scale_factor = 10**6
+scale_factor = config["scale_factor"]
 
 # the minimum area of a meet node (a node formed by the interestion of two disparate geograophic granularities), in square kilometers
-minimum_intersection_area = 1
+minimum_intersection_area = config["minimum_intersection_area"]
 
 # Define the abstract graph with a list of tuples with the form (source, destination), where source is a higher lower resolution granularity that encompasses destination, a higher resolution granularity.
-abstract_edges = [("usa48", "state"), ("state", "county"), ("usa48", "nerc"), ("usa48", "huc8"), ("usa48", "latlon")]
+abstract_edges = config["abstract_edges"]
 
 # Save the instance graph with its geometries included. This will create a very large graph.
-save_shapes = False
+save_shapes = config["save_shapes"]
 
 # open the shapefiles for each granularity
 states = read_file(f"{shapefile_dir}/state.shp").to_crs(epsg=projection)
@@ -257,7 +261,7 @@ print(metadata)
 
 # save the instance graph with its geometries (very large)
 if save_shapes:
-    with open("{}/{}-official-instance-graph-{}-shapes_{}km_simple1km.geojson".format(save_dir, "-".join(abstract_nodes), projection, minimum_intersection_area), mode='w') as outfile:
+    with open("{}/instance-graph_{}_{}_{}_{}_shapes.geojson".format(save_dir, "-".join(abstract_nodes), projection, minimum_intersection_area, config["tag"]), mode='w') as outfile:
         geojson.dump(json_graph.node_link_data(instance_graph), outfile)
 
 # remove geometries from the instance graph (much smaller)
@@ -267,9 +271,9 @@ for node in instance_graph_noshapes['nodes']:
         del node['shape']
 
 # save graphs to JSON files
-with open("{}/abstract-graph_{}_{}_{}km_simple1km.geojson".format(save_dir, "-".join(abstract_nodes), projection, minimum_intersection_area), mode='w') as outfile:
+with open("{}/abstract-graph_{}_{}_{}_{}.geojson".format(save_dir, "-".join(abstract_nodes), projection, minimum_intersection_area, config["tag"]), mode='w') as outfile:
     geojson.dump(json_graph.node_link_data(abstract_graph), outfile)
-with open("{}/instance-graph_{}_{}_{}km_simple1km.geojson".format(save_dir, "-".join(abstract_nodes), projection, minimum_intersection_area), mode='w') as outfile:
+with open("{}/instance-graph_{}_{}_{}_{}.geojson".format(save_dir, "-".join(abstract_nodes), projection, minimum_intersection_area, config["tag"]), mode='w') as outfile:
     geojson.dump(instance_graph_noshapes, outfile)
 
 print("done building graphs")
diff --git a/graphs/config.json b/graphs/config.json
new file mode 100644
index 0000000..3b40971
--- /dev/null
+++ b/graphs/config.json
@@ -0,0 +1,19 @@
+{
+    "projection": 3085,
+
+    "scale_factor": 1000000,
+
+    "minimum_intersection_area": 1,
+
+    "abstract_edges": [
+        ["usa48", "state"],
+	["state", "county"],
+	["usa48", "nerc"],
+	["usa48", "huc8"],
+	["usa48", "latlon"]
+    ],
+
+    "save_shapes": false,
+
+    "tag": "latest"
+}