set ipython's default scientific notation threshold

Question

How can I modify the point at which python decides to print in scientific notation?

E.g. I\'d like everything > 1e4 or < 1e-4 to be p

Answer 1

Actually it has nothing to do with IPython as it simply calls __str__. Sadly, float, being a built-in type, is implemented in C and you can't monkeypatch C types. So your only options are modifying __str__ implementation in cpython or patching IPython to be smarter than just calling __str__.

Answer 2

In IPython you can use

%precision %.4g

this will print floating point values who's absolute value is < 1e-4 or >= 1e4 in scientific notation.

You can find more information about the %precision command in the IPython API Docs. For string formatting options have a look at the Python Docs