A cheating sheet on Scipy, etc.

iPython

• ?+cmd = help(), help command
• %run python_cmd, run command externally
• %paste & %cpaste, block paste
• %timeit
• %reset
• %quickref / %magic, find all the magic cmd
• %hist, show cmd history
• %debug / %pdb, debug the code
• %xdel, del objects
• %logstart

Qt and Pylab

$ ipython qtconsole –pylab=inline # old feature $ ipython –pylab requires matplotlib and PyQt for inline plotting

shortcuts of command history

• Ctrl + P forward matching
• Ctrl + P backward matching
• Ctrl + P readline
• _iX: input in line X
• _X : output in line X

Interacting

• %cd / %pwd …
• %dirs / %alias
• !cmd

Bookmarks

• %bookmark markname DIRS
• %bookmark -l

Debug

    %debug
    u/d
    %pdb iX (X is variable)

Tricks

• reload after import (dreload: deep reload)
• import this

Numpy

Object

1. based on ndarray
2. basically, type in np is float64

• new: array(‘list’)/asarray(‘output’)
• zeros/ones(length): new a list of 0/1
• zeros/ones_like(list): new a list of 0/1 like given list
• eye/identity(n): diagonal matrix
• arange(length): new a number list

Type

dtype: variable type

astype(np.newtype) changing the type to newtype

Para

• ndim: dimension of array
• shape

Slice

changing of slice of an array will be broadcast

for copying, using array[:].copy()

index and slice, array[:,:,:]

x[1,2,…] is equivalent to x[1,2,:,:,:], x[…,3] to x[:,:,:,:,3] and x[4,…,5,:] to x[4,:,:,5,:]

Boolean index

1. array ==/!=/ value giving a boolean list with the same length, here the express array |/& value is used as boolean index.

2. array(boolean_list) selecting the true element and broadcasting.

data[data<0] = 0 # changing the negative elements to 0s

Fancy index

a([i:],[j:],[]…)

Function

• .reshape

• .T

• ufunc
  • a
• 2ufunc

b = np.fromfunction(f,(5,4),dtype=int)

Shape

Stack

• np.vstack/hstack
• newaxis
• np.r_

Split

1. No copy
2. Shallow copy
3. Deep copy

TBD `````````````````````````````````````

• python