C4V Hackathon - Born in Venezuela
Our project addresses the question:
can I go to my local hospital for xyz?. We created a process that takes the survey data and answers that question for 17 types of cases. We also created telegram bot that allows individuals to text and receive information about their local hospitals (details below).
We leverage Bigquery to pull data from the
angostura_dev.eh_health_survey_response and insert it into our own table
event-pipeline.hulthack.dashboard_v1. This data went into Tableau dashboard in the folder
tableau_sample/. The code is automated and any user can run the following function
code/make_data.py daily. There are no parameters needed in this code (however, the read/write permissions may need to be changed).
For the sample dashboard, you can run the code
sh setup.sh. We use dash to create a sample dashboard but the interactivity does not work and the text must be mapped to colors/unicode text.
For the sample bot, you can run the code
code/bot.py and the application will be live with the name
c4v_bot. The commands are
empieza with the name of the
state every hospital with all the main information.
To load the notebooks, type
jupyter lab --no-browser --port=3000 locally.
Data Quality Errors
Details can be found here:
C4V_Data_Quality_Issues.pdf. For example, the column
wahs_failure_sx is mispelt, and currently the data does not match this. Similarly the columns have differing definitions for not existing/not available ('No operativa', 'No existe', 'No hubo','Nunca ha existido').