

In [1]:

    
import numpy as np



In [3]:

    
import StringIO

Q1:

Saving a table to text with a header with no preceding "#"

Also, demo StringIO



In [15]:

    
x = StringIO.StringIO()
arr = np.arange(10)
np.savetxt(x,arr, header='test', comments="")



In [17]:

    
x.seek(0)
print(x.read())
with open('file.txt','w') as f:
    f.write(x.getvalue())









    



test
0.000000000000000000e+00
1.000000000000000000e+00
2.000000000000000000e+00
3.000000000000000000e+00
4.000000000000000000e+00
5.000000000000000000e+00
6.000000000000000000e+00
7.000000000000000000e+00
8.000000000000000000e+00
9.000000000000000000e+00



In [19]:

    
%%bash
cat file.txt









    



test
0.000000000000000000e+00
1.000000000000000000e+00
2.000000000000000000e+00
3.000000000000000000e+00
4.000000000000000000e+00
5.000000000000000000e+00
6.000000000000000000e+00
7.000000000000000000e+00
8.000000000000000000e+00
9.000000000000000000e+00

astropy convolution

How do you convolve fast?

see, e.g., http://keflavich.github.io/blog/fft-comparisons-in-python.html



In [20]:

    
from astropy.convolution import convolve, convolve_fft

Speed of DFT: $O(n^2)$

Speed of FFT: $O(n log(n))$



In [29]:

    
import scipy.fftpack, scipy.ndimage, scipy.signal



In [33]:

    
scipy.ndimage.convolve
scipy.signal.fftconvolve??



In [38]:

    
%%bash
factor 9216









    



9216: 2 2 2 2 2 2 2 2 2 2 3 3

faster fftw: --enable-avx for "advanced vector instructions". 8x FLOPs at a time!

What does it mean to "remove" fft modes?



In [72]:

    
x = np.random.randn(64) + 5 + np.sin(np.arange(64))*3



In [73]:

    
f = np.fft.fft(x)



In [74]:

    
pl.plot(x)









    Out[74]:





[<matplotlib.lines.Line2D at 0x10d470650>]



In [75]:

    
%matplotlib inline
import pylab as pl
pl.plot(np.abs(f))









    Out[75]:





[<matplotlib.lines.Line2D at 0x10e0ea110>]



In [76]:

    
f[0] = 0
f[10] = 0
f[64-10] = 0
pl.plot(np.abs(f))









    Out[76]:





[<matplotlib.lines.Line2D at 0x10e0e03d0>]



In [77]:

    
xi = np.fft.ifft(f)



In [78]:

    
pl.plot(xi.real)
pl.plot(x)









    Out[78]:





[<matplotlib.lines.Line2D at 0x10e3ec5d0>]

Q3

Install a module, then keep editing it.

python setup.py develop

Use https://github.com/astropy/package-template to get everything set up in a cool way. develop doesn't do much good for C code.

Within an interactive session, use reload(package) (python2) or import importlib; importlib.reload(package) to reload the package. This finnicky.

Other option, which works with C extensions: python setup.py build_ext --inplace. Or you can use python setup.py build to build into the build/ directory, which will then be accessible using import if you've used python setup.py develop



In [80]:

    
%%bash
cd ~/repos/astropy
ls
ls build/









    



CHANGES.rst
CITATION
CONTRIBUTING.md
MANIFEST.in
README.rst
ah_bootstrap.py
appveyor.yml
astropy
astropy.egg-info
astropy_helpers
build
cextern
dist
docs
ez_setup.py
licenses
pip-requirements
setup.cfg
setup.py
static
tox.ini
bdist.macosx-10.5-x86_64
lib.macosx-10.5-x86_64-2.7
lib.macosx-10.5-x86_64-3.5
temp.macosx-10.5-x86_64-2.7
temp.macosx-10.5-x86_64-3.5

last note

Profiling (tracking memory and time use) can be done with the memory_profile and timeit utilities. psrecord is also nice.



In [ ]: