pandem-source 1.0.9

Routinely collecting heterogeneous surveillance data into a coherent database

PANDEM - Source

Manage heterogeneous data sources for pandem 2 project.

Description

Identify, map and integrate multiple pandemic related data into a coherent pandemic-management database. Developed within the H2020 project PANDEM-2, Pandem-Source allows users to systematically capture, standardize and analyze data coming from international and national surveillance databases, participatory surveillance projects, social networks and mass media. This tool is focused on flexibility so adding new sources or variables can be easily done as is required during a pandemic episode.

Approach

Data integration follow a semantic approach. Data sources just need to be described using a Data Labelling Schema (DLS) file that specifies the acquisition chennel (URL, git, local file or script) the format and how to map the input data to Pandem-Source variables.

The DLS ensures each source and variable is properly documented knwing its origin, meaning and data quality.

Supported sources

Pandem-Source is designed to be flexible and extensible so new sources can be easily added by end users. In order to demontstrate this principle. The following sources are supported Out of the Box

Sources for indicators

• COVID19 Data Hub see source definition
• ECDC Atlas see source definition
• ECDC COVID19 Datasets see source definitions for variants, age group, daily cases, goverments measures, vaccination
• ECDC COVID19 Simulated data see source definition
• Serotracker see source definition
• Open Sky Nerwork see source definition
• Influenza Net see source definition
• MediSys see source definition
• Twitter see source definition
• User provided data. You can use this data template to easily integrate your own real or ficticious datasets.

Sources for referentials

• Eurostats NUTS see source definition
• Our airports see source definition
• Geonames see source definition
• ICD-10-CM see source definition

User Interface - timeseries explorer

User Interface - Integration dashboard

Requirements

• Python 3.7 or higher
• R 3.6.3 or higher
• Docker (optional for supporting machine learning article classification)

Installation

pip install pandem-source

Running Pandem-Source

• Set the PANDEM_HOME variable to a local foder

export PANDEM_HOME=your data folder here 

• Load default sources

python -m pandemsource reset --restore-factory-defaults

• Running monitoring and dashboard

#If you do not have docker or NLP models you have to run pandem WITHOUT NLP without Twitter and without MediSys
python -m pandemsource start -d --no-nlp  
#If you have docker and NLP models the you can run without limitations
python -m pandemsource start -d 

• Accessing the dashboard from http://localhost:8001 to see the progress and integrated time series

Troubleshooting

If the command python -m pandemsource start -d --no-nlp does not work. Please make sure you've installed R 3.6.3 or higher. If R is properly installed, please make sure you've also installed required dependencies with:

(In your terminal)

sudo apt install libxml2-dev libsodium-dev libssl-dev libcurl4-openssl-dev

(In the R interpreter)

install.packages(c("epitweetr", "dplyr", "shiny", "plotly", "DT", "jsonlite", "httr"))

PANDEM 2

PANDEM-2 is a H2020 EU-funded project that aims to develop new solutions for efficient, EU-wide pandemic management. The goal of PANDEM-2 is to prepare Europe for future pandemics through innovations in training and to build capacity between EU member states responding to pandemics on a cross-border basis.