Announcing Deedle – Data Frame and Time Series Package for Exploratory Data Programming with F# and C#
We’re very pleased to announce the availability of the Deedle package for Data Frame and Time Series programming with F# and C#.
Deedle is an open library developed by the F# community with the major contributors being BlueMountain Capital. Here’s an excerpt from the announcement by BlueMountain Capital yesterday:
Yesterday we announced Deedle, our new Open Source library for exploratory data analysis in C# and F#. Deedle (almost) stands for “Dotnet Exploratory Data Library”. This is a library with similar (but not yet as broad) capabilities as the widely respected Pandas library for Python.
Deedle was developed by Tomas Petricek, with assistance from Adam Klein and myself.
We are finding Deedle to be extremely powerful for research. We hope others will find it similarly useful and make improvements to make it an even better package.
Deedle has great documentation and examples so we hope you will get up to speed quickly! A video of the talk and the sample source code will be posted ASAP.
Data Frame programming is fast becoming the “standard” way of doing multi-dimensional and statistical data processing in systems such as R, Python and now .NET. If you’ve never done data frame or time series programming, I strongly recommend you learn this space by working with Deedle. The DataFrame paradigm is an awesomely paradigm for indexed data processing and statistical analysis, including the crucial and growing area of time-indexed data. Deedle includes an API explicitly designed for use from F#, and great documentation is available for F# users.
Deedle is also an open library with a license that permits broad use, meaning you can contribute to it through contributions on GitHub. Deedle is also available as a NuGet package and can be used on OSX and Linux.
Deedle is a significant part of the growing ecosystem of F# tools and libraries for data-science and should be considered the “canonical” way of doing data frame programming in F#. It is also a strong contender for being the best way to do data frame programming on .NET and with C#.
The Visual F# Tools Team