Quandl, a company that launched a search engine for numerical data in January 2013, plans on introducing a brand new uploading tool in early 2014 which will make it possible for users to upload their own datasets to the Quandl platform. In February of this year, the Toronto-based company launched the Quandl API which provides programmatic access to every single dataset available on Quandl.
The collaboratively-curated portal, started out with 2 million financial, economic and sociological datasets. In November, the number of available datasets on Quandl reached 8,000,000 and the number continues to grow rapidly. Quandl features time-series data in a variety of domains including finance, economics, demography, sociology, markets, energy, health, education and the environment.
Quandl graphs can be embedded on websites, blogs and other applications. A graph of any dataset on Quandl, including supersets, can be embedded.
The Quandl API allows developers to programmatically access all datasets on Quandl regardless of how, where, and in what format the data was originally published. ProgrammableWeb’s Ajay Ohri wrote an excellent article that includes some examples of how to use the Quandl API. The availability of the Quandl API has helped to increase the number of Quandl users and allowed developers to create innovative applications driven by the Quandl platform. Tammer Kamel, Quandl CEO and Founder, told ProgrammableWeb that “The API is absolutely fundamental to our value proposition for a very substantial fraction of our user base.”
Tammer Kamel also described the three ways the API is being used:
Most of Quandl’s datasets are univariate making them ideal to use with the BigML platform
One great example of a platform using the Quandl API is BigML, a cloud based machine learning platform that allows users to quickly build predictive models with Quandl datasets and Supersets. In a BigML blog post, Candido Zuriaga, Lead Data Wrangler, explains why Quandl works well with BigML to create “powerful financial, economic, and social predictive models”:
“Most of Quandl’s datasets are univariate, which provide interesting insight and lend themselves to interesting time-series forecasting models. In addition, there’s a great utility in which you can combine columns from different datasets to create a more complex item called a Superset. BigML works very well with these multivariate Supersets.”
Earlier this month, DataHero announced a new partnership with Quandl allowing sample data sets to be provided to DataHero users. DataHero is a fairly new data visualization and analysis platform that allows users to import, visualize and analyze data without the need to know code or have extensive technical knowledge. DataHero is using the sample datasets to help new users understand how to use the DataHero platform.
Other examples of applications and platforms that incorporate Quandl data include in-browser algorithmic trading platform Quantopian, R language-based statistical application Rapporter, and mathematical and analytical software Maplesoft.
Tammer Kamel told ProgrammableWeb that the primary goal of Quandl is that it should be as easy to “put data on Quandl as it is to get data from Quandl.” The new uploading tool coming in early 2014 will help Quandl to achieve this goal. He went on to explain that:
“In 2013, our focus was on building an API for the consumption of Quandl data. While we will continue to refine that API, our attention is now turning to creating an API–indeed an entire PaaS–for users to put data on Quandl.”
For more information about the Quandl platform and the Quandl API, visit the official website.
By Janet Wagner. Janet is a Data Journalist and Full Stack Developer based in Toledo, Ohio. Her focus revolves around APIs, open data, data visualization, and data-driven journalism. Follow her on Twitter, Google+, and LinkedIn.