Pandas -- not just for data scientists (Uzi Halaby Senerman, BlueVine)

• The slowness of Python comes from its dynamic nature, not the fact that it's interpreted.
• NumPy and Pandas use statically typed arrays and vector functions.
• Pandas data frames are like Excel tables and pandas gives us an API for working with data in them.
• If you use Jupiter, there are also GUI-ish components.
• You can apply regular functions to data via apply but that's much slower than using vector functions (ufuncs) and should generally be avoided.
• Thinking in SQL-like terms (instead of iterating over rows) and using pandas accordingly often produces much better performance.

References

• Python for Data Analysis (http://shop.oreilly.com/product/0636920023784.do)