|
Size: 3944
Comment:
|
← Revision 5 as of 2008-10-30 13:08:53 ⇥
Size: 3280
Comment:
|
| Deletions are marked like this. | Additions are marked like this. |
| Line 98: | Line 98: |
| === Importing modules === There are many ways to import modules. The resulting namespaces are different * Import module SciPy: {{{import scipy}}} * List SciPy subpackages {{{help(scipy)}}} or {{{scipy.info(scipy)}}} * Import SciPy subpackage n-dimensional image package {{{import scipy.ndimage}}} * List ndimage functions {{{help(scipy.ndimage)}}} * Help on ndimage function laplace {{{help(scipy.ndimage.laplace)}}} * Import SciPy module into local namespace {{{from scipy import *}}} * Import SciPy subpackage ndimage in namespace ndi {{{import scipy.ndimage as ndi}}} * Write source for this object to output {{{scipy.source(ndi.laplace)}}} |
This lesson motivates the use of Python. The Python programming language is compared with other systems for satellite image processing and analysis.
Software for Satellite Image processing and analysis
Comparison of Programming Languages
Ideal programming language for satellite image processing and analysis
- Fast array operations
- Image processing and numeric/scientific routines
- Visualization
- Various data formats
- Processing of files and metadata
- Short development cycles
- Very high level of abstraction
- Interactive
Ideal programming language for satellite image processing and analysis?
- Assembler
- Fortran, C/C++, Java
- Perl, Python
- Matlab, IDL
- Visual and menu driven environments, ENVI, GIS
Programming versus visual environments
- Visual environments are very useful for specific tasks
- Closed commerical software
- Programming offers more flexibility
- ESRI's ArcGIS scripting with python
Scripting verus Traditional Programming
- Traditional programming refers to building usually large, monolithic systems
- Fortran, C/C++, Java
- Scripting means programming at a high and flexible abstraction level
- Perl, Python, Ruby, Scheme, Tcl
- Scientific computing environments
- IDL, Matlab/Octave, Maple, Mathematica, R
Why Python?
Scalability
- The ability to scale from easy to difficult problems and the ability for beginners and experts to be comfortable.
- Python is easy enough to be a first language and powerful enough to write complex applications
- Scientific computing is more than number crunching
- Converting data formats
- Extracting metadata from text
- Working with a large number of files and directories
- Object oriented programming possible, but not required
- Freely available and runs on Unix, Mac and Windows
- Simple interfacing of C,C++ and Fortran code
- Heterogeneous data structures are easy to use
- Readable and compact code
Scientific Python Environment
- Basic Python has limited instruction set
- Extensions (modules)
- Scientific modules (scipy/numpy)
- Interactive environment (ipython)
- Plots and visualization (pylab)
Getting started with IPython
Invoking the IPython shell
ipython ipython -pylab
loads pylab module and enables interactive plotting
Quit with CTRL-D
Getting help
help() help modules
list available modules
Features of IPython:
- Command history (up, down)
- Completion by typing TAB
- System commands through magic functions cd, ls, env, pwd
- Access to Unix shell by prefix ! {!ls -s $|$ sort -g}
- Debugging and profiling
Program control: run
Print interactive variables who, whos
Defining functions and visualization
def tv(a):
imshow(a,interpolation='nearest')
colorbar()
a=rand(10,10)
tv(a)
savefig('figure1.png',dpi=100)x=linspace(0,2*pi,100)
plot(x,sin(x),x,cos(x))
grid()
axis('tight')
legend(('sin(x)','cos(x)'),'upper right')
xlabel('x')
ylabel('f(x)')