(i) Number one feature: system wide symbol searching! When ipython is loaded, suggestions module will scan your python paths for symbols and will create a cache. On the author's old 2008 computer (this readme is from 2017), 50000 symbols load in 10 seconds. But then you get a very easy way to import any symbol, without typing or remembering the entrine import path. First example:

In [1]: %findsymbol DecisionTreeClasifir # two typos here on purpose Out[1]: 0 (C) from sklearn.tree import DecisionTreeClassifier

In [2]: %suggestion 0 from sklearn.tree import DecisionTreeClassifier # it's now imported!

%findsymbol searches string up to two character edits (deletion, substitution, transpose and insertion).

Second example:

In [1]: %findsymbol pypl  # now hit tab!
[this completes to:]
In [1]: %findsymbol pyplot...matplotlib  # press enter now.
from matplotlib import pyplot
[pyplot is now imported]

Even better example:

In [1]: %findsymbol pypl  # now hit tab!
[this completes to:]
In [1]: %findsymbol pyplot...matplotlib
[now add -as parameter]
In [1]: %findsymbol pyplot...matplotlib -as plt
from matplotlib import pyplot as plt
[pyplot is now imported as plt]

The completions offered by pressing tab in a %findsymbol line are all the symbols that begin with what you wrote. Note that this is case-sensitive.

This also works in jupyter :)

(ii) Get suggestions on misspelled names:

In [1]: my_awesome_variable = 10

NameError Traceback (most recent call last) in () ----> 1 10 * my_awsome_variable ** 3

NameError: name 'my_awsome_variable' is not defined Did you mean: 0 my_awesome_variable

In [3]: %suggestion 0 [ipython automatically fills the next line] In [4]: 10 * my_awesome_variable ** 3

Auto-filling of corrected code currently only works inside the shell and not in jupyter.

Run:

pip install git+https://github.com/drorspei/ipython-suggestions # from Github

then append the output of ipython -m ipython_suggestions to the output of ipython profile locate (typically ~/.ipython/profile_default/ipython_config.py).