Joint TBDSG/TBPy Meeting: Data Wrangling with Pandas Dataframe PLUS...

LAST MINUTE VENUE CHANGE!!!!!
Due to a desire to maintain marital bliss, our meeting will now be at the Undercroft, (1326 E 9th Ave, Tampa) you may know it as the old SOFWERX location a block or so south of Palm Ave in Ybor.

https://www.theundercroft.net/

Sharp Decisions will be providing food and beverages. They'll appreciate if if you keep your RSVP current :)

-----------------------------------------------------------------------
Tampa Bay Data Scientists!

Join us for a joint meeting of TBDSG and TBPy.
https://www.meetup.com/tampa-bay-python

This session will review an implementation of the Python pandas library as a data transformation and aggregation layer between a Postgres database and an Elastic Kibana data visualization front-end that is hosted entirely in the cloud on AWS. The business case was to securely transform electronic
medical imaging metadata into visual metrics to identify which imaging clinics were optimizing the capabilities of a medical records solution. The
highlights of the implementation that will be explored include:

• how database cursor techniques were necessary to throttle the dataload in order to obey AWS memory constraints,
• the efficient capabilities of the pandas dataframe to load and cache large volumes of data,
• a look at how successive groupby operations transformed the data into weekly, monthly, quarterly and annual perspectives of trends,
• usage of the merging capabilities to mix data from separate tables into one dataframe,
• and finally to efficiently export data in JSON format to a NoSql Elasticsearch database where the Kibana visualization system could render the insights.

About Our Speaker:

JP Antona combines experience of application development technologies (Java, Python, Javascript) and data analytics engineering. He is responsible for engineering of Salesforce's Einstein Prediction Builder product which provides generalized binary classification and regression machine learning capabilities to any standard or customized data entity on the Salesforce. His focus is on the user interface and business logic components of Prediction Builder as well as the interface to the TransmorgifAI -based machine learning system.

HOW TO FIND US:
Not so tricky as long as you see THAT THE VENUE HAS CHANGED. SE corner of building.

Try street parking or the municipal garage to the west. Easier to get to, but at a price.