opendrr-api

REST API for OpenDRR data

opendrr-api is a repository used in conjunction with the model-factory repository. It contains Python, Bash, and Dockerfile scripts to successfully implement a REST API for OpenDRR data which includes a PostGIS database, Elasticsearch, Kibana, and pygeoapi. The model-factory repository contains the necessary scripts to transform the OpenDRR source data into risk profile indicators that go into the PostGIS database.

• postgis/
  • Scripts to setup the postGIS database.
• pygeoapi/
  • Scripts to setup pygeoapi.
• python/
  • Scripts to process opendrr source data, model-factory scripts, and load elasticsearch indexes.
• docker-compose-run.yml, docker-compose.yml, opendrr-config_template.yml
  • docker compose and opendrr config files
• requirements.txt
  • list of modules and versions required to be installed. $ pip install -r requirements.txt

Refer to the releases section for latest version changes.

How to build your own stack - Setup in your local environment

1. Prerequisites

• Docker engine installed and running
• Download or clone this repository to your local development environment

2. Edit the Docker environment settings

Make a copy of the sample.env file and rename it to .env. Make changes if required otherwise leave the default settings.

The settings below can be found in the .env file and can be adjusted to 'true' or 'false' depending on your preference. For example, if you want to load the PSRA data into their own PostGIS database, Elasticsearch, and Kibana, you can set processPSRA and loadPsraModels to 'true' and have all other options set to 'false'. Specifying the features that are only required can save you time.

Processing the Earthquake Scenarios (DSRA) / Probabilistic Earthquake Risk (PSRA) source data:

processDSRA=true
processPSRA=true

Loading the indexes into Elasticsearch and Kibana:

loadDsraScenario=true
loadPsraModels=true
loadHazardThreat=false
loadPhysicalExposure=true
loadRiskDynamics=true
loadSocialFabric=true

3. Edit the Python container configuration

Make a copy of python/sample_config.ini and rename it config.ini. Open this file in an editor, add the required github_token and set the remaining parameters as follows:

[auth]
# GitHub Token for Private Repo Accesss
github_token = ghp_XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

[rds]
# PostGIS Connection Details
postgres_host = db-opendrr
postgres_port = 5432
postgres_un = postgres
postgres_pw = password
postgres_db = opendrr
postgres_address = db-opendrr:5432/opendrr

[es]
# Elasticsearch Connection Details
es_un = elastic
es_pw = changeme
es_endpoint = elasticsearch-opendrr:9200
kibana_endpoint = localhost:5601
# version of indices to be configured (i.e. v1.4.1)
index_version = v1.4.4

4. Run docker-compose

docker-compose up --build

NOTE: you will see errors thrown by the opendrr-api_pygeoapi-opendrr_1 container as the stack builds. These can be ignored.

Once the stack is built the user will need to verify that everything is working.

NOTE: you can stop the stack whenever you like with Ctrl-C or docker-compose stop. See below on how you can bring the stack back up without re-building.

5. Verify that everything is working

Check Elasticsearch to ensure that the indexes were created:

http://localhost:9200/_cat/indices?v&pretty

You should see something similar to:

health status index ...
green  open   afm7p2_lrdmf_scenario_shakemap_intensity_building
green  open   .apm-custom-link
green  open   afm7p2_lrdmf_damage_state_building
green  open   .kibana_task_manager_1
green  open   afm7p2_lrdmf_social_disruption_building
green  open   .apm-agent-configuration
green  open   afm7p2_lrdmf_recovery_time_building
green  open   afm7p2_lrdmf_scenario_shakemap_intensity_building
green  open   afm7p2_lrdmf_casualties_building
green  open   .kibana_1

You can explore the indexes in Elasticsearch using Kibana

http://localhost:5601

Check pygeoapi to make sure the collections can be accessed

http://localhost:5001/collections

Feature collections can be accessed as follows or by clicking on the links provided on the collections page

http://localhost:5001/collections/afm7p2_lrdmf_scenario_shakemap_intensity_building/items?f=json&limit=1

You should see something similar to:

{
    "type": "FeatureCollection",
    "features": [
        {
        "type": "Feature",
        "geometry": {
            "type": "Point",
            "coordinates": [
            -117.58484079,
            49.58943143
            ]
        },
        "properties": {
            "AssetID": "59002092-RES3A-W1-PC",
            "Sauid": "59002092",
            "sL_Asset_b0": 27.602575,
            "sL_Bldg_b0": 27.602575,
            "sLr_Bldg_b0": 1,
            "sLr2_BCR_b0": 0.000819,
            "SLr2_RoI": 0.001359,
            "sL_Str_b0": 27.602575,
            "sLsd_Str_b0": 23.4638,
            "sL_NStr_b0": 0,
            "sLsd_NStr_b0": 0,
            "sL_Cont_b0": 0,
            "sLsd_Cont_b0": 0,
            "sL_Asset_r2": 0.311704,
            "sL_Bldg_r2": 0.311704,
            "sLr_Bldg_r2": 1,
            "sL_Str_r2": 0.311704,
            "sLsd_Str_r2": 0.264966,
            "sL_NStr_r2": 0,
            "sLsd_NStr_r2": 0,
            "sL_Cont_r2": 0,
            "sLsd_Cont_r2": 0,
            "geom_point": "0101000020E6100000AD9A10086E655DC0D18A357D72CB4840"
        },
        "id": "59002092"
        }
    ],
    "numberMatched": 173630,
    "numberReturned": 1,
    "links": [
        {
        "type": "application/geo+json",
        "rel": "self",
        "title": "This document as GeoJSON",
        "href": "http://localhost:5001/collections/afm7p2_lrdmf_scenario_shakemap_intensity_building/items?f=json&limit=1"
        },
        {
        "rel": "alternate",
        "type": "application/ld+json",
        "title": "This document as RDF (JSON-LD)",
        "href": "http://localhost:5001/collections/afm7p2_lrdmf_scenario_shakemap_intensity_building/items?f=jsonld&limit=1"
        },
        {
        "type": "text/html",
        "rel": "alternate",
        "title": "This document as HTML",
        "href": "http://localhost:5001/collections/afm7p2_lrdmf_scenario_shakemap_intensity_building/items?f=html&limit=1"
        },
        {
        "type": "application/geo+json",
        "rel": "next",
        "title": "items (next)",
        "href": "http://localhost:5001/collections/afm7p2_lrdmf_scenario_shakemap_intensity_building/items?startindex=1&limit=1"
        },
        {
        "type": "application/json",
        "title": "Economic loss buildings",
        "rel": "collection",
        "href": "http://localhost:5001/collections/afm7p2_lrdmf_scenario_shakemap_intensity_building"
        }
    ],
    "timeStamp": "2020-08-18T22:46:10.513010Z"
    }

Interacting with the endpoints

Querying pygeoapi

Refer to the pygeoapi documentation for general guidance:

http://localhost:5001/openapi?f=html

`

NOTE: querying is currently limited to spatial extent and exact value queries. For more complex querying use Elasticsearch (see below).

To filter on a specfic attribute

http://localhost:5001/collections/afm7p2_lrdmf_scenario_shakemap_intensity_building/items?sH_Mag=7.2

To filter using a bounding box

http://localhost:5001/collections/afm7p2_lrdmf_scenario_shakemap_intensity_building/items?bbox=-119,48.8,-118.9,49.8&f=json

Querying Elasticsearch

Range query

http://localhost:9200/afm7p2_lrdmf_scenario_shakemap_intensity_building/_search?q=properties.sH_PGA:[0.047580+TO+0.047584]

OR using curl:

curl -XGET "http://localhost:9200/afm7p2_lrdmf_scenario_shakemap_intensity_building/_search" -H 'Content-Type:
application/json' -d'
{
    "query": {
        "range": {
            "properties.sH_PGA": {
                "gte": 0.047580,
                "lte": 0.047584
            }
        }
    }
}'

Specific value

http://localhost:9200/afm7p2_lrdmf_scenario_shakemap_intensity_building/_search?q=properties.sH_PGA:0.047584

OR using curl:

curl -XGET "http://localhost:9200/afm7p2_lrdmf_scenario_shakemap_intensity_building/_search" -H 'Content-Type:
application/json' -d'
{
    "query": {
        "match": {
            "properties.sH_PGA" : 0.047584
        }
    }
}'

Bounding box query

curl -XGET "http://localhost:9200/afm7p2_lrdmf_scenario_shakemap_intensity_building/_search" -H 'Content-Type:
application/json' -d'
{
    "query": {
        "bool": {
            "filter": [
                {
                    "geo_shape": {
                        "geometry": {
                            "shape": {
                                "type": "envelope",
                                "coordinates": [ [ -118.7, 50 ], [ -118.4, 49.9 ] ]
                            },
                            "relation": "intersects"
                        }
                    }
                }
            ]
        }
    }
}'

Nearest query

curl -XGET "http://localhost:9200/nhsl_hazard_threat_all_indicators_s/_search" -H 'Content-Type:
application/json' -d'
{
  "query": {
    "geo_shape": {
      "geometry": {
        "shape": {
          "type": "circle",
          "radius": "20km",
          "coordinates": [ -118, 49 ]
        }
      }
    }
  }
}'

Interacting with the spatial database

The spatial database is implemented using PostGIS. You can connect to PostGIS using pgAdmin with the connection parameters in your .env file. For example:

POSTGRES_USER: postgres
POSTGRES_PASSWORD: password
POSTGRES_PORT: 5432
DB_NAME: opendrr

Adding datasets to QGIS

You have two options:

Connect to PostGIS

1. Add a "New Connection" by right clicking on the "PostGIS" data type in the browser
2. Enter a name for your connection (i.e. "OpenDRR")
3. Add the credentials as per your .env file (see above)
4. Click the "OK" button

Connect to OGC OpenAPI - Features

1. Add a "New Connection" by right clicking on the "WFS / OGC API -Features" data type in the browser
2. Enter a name for your connection (i.e. "OpenDRR")
3. Enter http://localhost:5001 in the URL field
4. Select "OGC API - Features" in the "Version" dropdown
5. Click the "OK" button

Start/Stop the stack

Once the stack is built you only need to re-build when there is new data. The docker-compose-run.yml script is an override that you can use to run the built stack - it doesn't create the python container that pulls the latest code and data from GitHub to populate the stack.

To start the stack:

docker-compose -f docker-compose-run.yml start

To stop the stack:

docker-compose -f docker-compose-run.yml stop

Updating or rebuilding the stack

Take the stack down and remove the volumes:

docker-compose down -v

Rebuild the stack:

docker-compose up --build