Continuum Analytics, the creator and driving force behind Anaconda, the leading Open Data Science platform powered by Python, announced the availability of Anaconda Mosaic. With the ability to easily create and catalog transformations against heterogeneous data stores, Anaconda Mosaic empowers data scientists, quants and business analysts to interactively explore, visualize, and transform larger-than-memory data sets to more quickly discover new insights.
Enterprise data architecture is becoming increasingly complex. Data stores have a relatively short half life and data is being shifted to new data stores – NoSQL, SQL, flat files – at a higher frequency. In order for organizations to find insights from the data they must first find existing transformations and rewrite the transformations for the new data store. This creates delays in getting the insights from the data. Continuum Analytics’ Anaconda Mosaic enables organizations to quickly explore, visualize, and redeploy transformations based on Pandas and SQL without rewriting the transformations while maintaining governance by tracking data lineage and provenance.
Through the course of daily operations, businesses accumulate huge amounts of data that gets locked away in legacy databases and flat file repositories. The transformations that made the data usable for analysis gets lost, buried or simply forgotten,” said Michele Chambers, Executive Vice President Anaconda Business Unit & CMO at Continuum Analytics. “Our mission is for Anaconda Mosaic to unlock the mystery of this dark data, making it accessible for businesses to quickly redeploy to new data stores without any refactoring so enterprises can reap the analytic insight and value almost instantly. By eliminating refactoring of transformations, enterprises dramatically speed up their time-to-value, without having to spend lengthy cycles on the refactoring process.”
Some of the key features of Anaconda Mosaic include:
- Visually explore your data. Mosaic provides built-in visualizations for large heterogeneous datasets that makes it easy for data scientists and business analysts to accurately understand data including anomalies.
- Instantly get portable transformations. Create transformations with the expression builder to catalog data sources and transformations. Execute the transformation against heterogeneous data stores while tracking data lineage and provenance. When data stores changes, simply deploy the transformations and quickly get the data transformed and ready for analysis.
- Write once, compute anywhere. For maximum efficiency, Mosaic translates transformations and orchestrates computation execution on the data backend, minimizing the costly movement of data across the network and taking full advantage of the built-in highly optimized code featured in the data backend. Users can access data in multiple data stores with the same code without rewriting queries or analytic pipelines.
- Harvest large flat file repositories in place: Mosaic combines flat files, adds derived data and filters for performance easily. This allows users to describe the structure of their data in large flat file repositories and uses that description in data discovery, visualization, and transformations, saving the user from writing tedious ETL code. Mosaic ensures that the data loaded is only what is necessary to compute the transformation, which can lead to significant memory and performance gains.
Sign up for the free insideBIGDATA newsletter.
Speak Your Mind