On DataFrame datatype in Ruby
TL;DR: We need DataFrame as a data structure in Ruby. There are several promising candidates but no one with good usability. Some considerations on requirements to good DataFrame library are proposed, alongside with some rants on using Ruby for science.
As programming languages evolve, our notion of “basic necessary” data types for high-level language also evolves. So, nowadays in modern languages we have different kinds of numbers for different tasks; we have arrays (which are not “just pointers to memory area”); we have strings (which are not “just arrays of characters”), hashes/dictionaries (which are not “some specialized algorithmic concept, available through separate library”), regexps, ranges and so on.
Having complex types in a language core and having literals for them has a great value: programmers don’t need to reinvent something already considered as a wheel. (If you tried to use together several C++ libraries, each having different classes for strings and arrays, you’ll understand.)
(“Standard” here does not mean adding to core language or even standard library—for now—but just well-designed, highly usable and widely aknowledged gem.)
