pyIDL 0.6

pyIDL
# Python bindings for IDL, based on Andrew McMurry's python-IDL
# Requires ITT's IDL and numarray
# Download pyIDL source tarfile
# Installation

unzup and unpack the gzipped tar archive
$ tar -xvzf pyIDL-0.6.tgz

set environment variables
(paths will depend on IDL installation and platform)
$ export IDL_VERSION = '6.4'
$ export IDL_INCDIR = '/itt/idl/external/include'
$ export IDL_BINDIR = '/itt/idl/bin/bin.linux.x86'
$ export X11_LIBDIR = '/usr/lib/' (linux only)

install the package
$ python setup.py build
$ python setup.py install

(alternatively, a Makefile is included)

# Documentation

import the idl class
>>> from pyIDL import idl

instantiate the idl class
>>> ri = idl()

get help
>>> print ri.doc()
Python-IDL bindings
eval(command):
execute an IDL command
get(name,[array,allowNone]):
fetch a variable from IDL
put(name,value,[array,type,allowNone]):
push a variable to IDL
who([name,local,stdout]):
print/return the IDL/local variables
help([name]):
print the IDL help message (for a variable)
delete(name):
destroy selected IDL variables
map([name]):
get the IDL data type mapping
prompt():
an interactive IDL session
_print(value):
print using the IDL print command

push python variables into IDL and interact with IDL scripting language
>>> from numarray import *
>>> x = []
>>> for i in range(21): x.append(i*pi/10)
...
>>> ri.put('x',x)
>>> ri.eval('y = sin(x)')
>>> ri.eval('plot, x,y')

use the interactive prompt
>>> ri.prompt()
IDL interface:
% At $MAIN$
X DOUBLE = Array[21]
Y DOUBLE = Array[21]
Compiled Procedures:
$MAIN$

Compiled Functions:

IDL> IF x[0] LT 1 $
IDL> THEN print, x
0.0000000 0.31415927 0.62831853 0.94247780 1.2566371
1.5707963 1.8849556 2.1991149 2.5132741 2.8274334
3.1415927 3.4557519 3.7699112 4.0840704 4.3982297
4.7123890 5.0265482 5.3407075 5.6548668 5.9690260
6.2831853
IDL> exit

check variables in IDL session
>>> ri.who().keys()
['Y', 'X']
>>> ri.who('y')
[0.0, 0.3090169943749474, 0.58778525229247314, 0.80901699437494745,
0.95105651629515353, 1.0, 0.95105651629515364, 0.80901699437494745,
0.58778525229247325, 0.30901699437494751, 1.2246063538223773e-16,
-0.3090169943749469, -0.58778525229247303, -0.80901699437494734,
-0.95105651629515353, -1.0, -0.95105651629515364, -0.80901699437494756,
-0.58778525229247336, -0.30901699437494762, -2.4492127076447545e-16]

get variables back into python from IDL
>>> sinx = ri.get('y')

use shortcuts for put, eval, and get
>>> ri.z = 0.5
>>> ri('print, z')
0.50000000
>>> ri.z + 1
1.5

delete variables from IDL
>>> ri.delete('x')
>>> ri.delete('y')

use IDL function/procedure directly from python
>>> ri.sin(x)
[0.0, 0.3090169943749474, 0.58778525229247314, 0.80901699437494745,
0.95105651629515353, 1.0, 0.95105651629515364, 0.80901699437494745,
0.58778525229247325, 0.30901699437494751, 1.2246063538223773e-16,
-0.3090169943749469, -0.58778525229247303, -0.80901699437494734,
-0.95105651629515353, -1.0, -0.95105651629515364, -0.80901699437494756,
-0.58778525229247336, -0.30901699437494762, -2.4492127076447545e-16]

# Versions

0.6: 07/18/07
added support for idl_6.4
fixed boolean support in direct IDL function/procedure calls
0.5: 04/18/07
fixed support for python2.5
0.4.1: 12/01/06
fixed support for mac os
0.4: 10/06/06
added support for idl_6.3
improved support for NumPy
removed dependency on Numeric
0.3: 06/09/06
improved support for Numeric
added _print
extended examples
0.2: 05/19/06
support for 64-bit platforms
shortcuts for put, get, eval
access IDL functions/procedures from python
added examples directory
0.1: 06/21/05
initial; python bindings for RSI's IDL
interactive IDL prompt