NumericPython - Revision history

GethinWilliams: /* Input and Output */

2009-07-22T18:26:34Z

Input and Output

2009-07-22T17:00:29Z

Visualising the Contents of an Array

2009-07-22T16:49:50Z

Visualising the Contents of an Array

2009-07-22T16:18:15Z

Visualising the Contents of an Array

2009-07-22T16:16:37Z

Interest

2009-07-22T16:10:55Z

Arrays in numpy

2009-07-22T16:01:26Z

Arrays in numpy

2009-07-22T15:26:17Z

2009-07-22T15:12:48Z

Some Higher-Level Features

2009-07-22T15:08:08Z

Arrays in numpy

← Older revision		Revision as of 18:26, 22 July 2009
Line 251:		Line 251:

	=Input and Output=		=Input and Output=
		+
		+	The foregoing is all very interesting, but life would be rather dull if you had to re-enter all your data by hand whenever you set to work with Python and numpy. Therefore we need a means to save data to a file and load it again. Happily, we can do this rather easily using a couple of routines from the '''pylab''' package:
		+
		+	<pre>
		+	>>> from numpy import *
		+	>>> from pylab import load
		+	>>> from pylab import save
		+	>>> data = zeros((3,3))
		+	>>> save('myfile.txt', data)
		+	>>> read_data = load("myfile.txt")
		+	</pre>
		+
		+	'''warning, the load() function of numpy will be shadowed''' in the above example. One way to protect yourself against this is to make use of '''namespaces''': Modify your import command to '''import pylab''' and then use '''pylab.load(..)'''.

← Older revision		Revision as of 17:00, 22 July 2009
Line 236:		Line 236:

	[[Image:Sinc-matplotlib-contour.png\|thumb\|300px\|none\|A contour map of the sinc function]]		[[Image:Sinc-matplotlib-contour.png\|thumb\|300px\|none\|A contour map of the sinc function]]
		+
		+	The '''clip''' function is an interesting one. Using it, we can replace values greater than some threshold with a ceiling and lower than another threshold with a floor. For example:
		+
		+	<pre>
		+	>>> Z1 = clip(Z,-0.1,0.1)
		+	>>> plt.figure()
		+	>>> plt.contour(X,Y,Z1)
		+	>>> plt.show()
		+	</pre>
		+
		+	gives us:
		+
		+	[[Image:Sinc-contour-truncated.png\|thumb\|300px\|none\|A contour map of the sinc function]]

	=Input and Output=		=Input and Output=

@@ Line 218: / Line 218: @@
 =Visualising the Contents of an Array=
-The above examples are quite natty, but we have deliberately kept the array sizes small so that we can print the element values easily.  In practice, you may find that your array sizes are much larger and printing the values to the screen is impractical.  Fear not!  Python has many packages which help you plot your data, so that you can explore it.  Two of many possibilities are outlines below, where we have crafted an array of values which are more pleasing to the eye thusly:
+The above examples are quite natty, but we have deliberately kept the array sizes small so that we can print the element values easily.  In practice, you may find that your array sizes are much larger and printing the values to the screen is impractical.  Fear not!  Python has many packages which help you plot your data, so that you can explore it.  One of the many possibilities are outlines below, where we have crafted an array of values which are more pleasing to the eye thusly:
 <pre>
@@ Line 225: / Line 225: @@
 >>> z1 = sqrt(add.outer(x**2,y**2))
 >>> Z = sin(z1)/z1
 >>> import matplotlib.pyplot as plt
 >>> from pylab import meshgrid

@@ Line 221: / Line 221: @@
 <pre>
-x = arange(-5,10)
+>>> x = arange(-5,10)
-y = arange(-4,11)
+>>> y = arange(-4,11)
-z1 = sqrt(add.outer(x**2,y**2))
+>>> z1 = sqrt(add.outer(x**2,y**2))
-Z = sin(z1)/z1
+>>> Z = sin(z1)/z1
 </pre>
@@ Line 230: / Line 230: @@
 <pre>
-from NumTut import *
+>>> from NumTut import *
-view(Z)
+>>> view(Z)
 </pre>
@@ Line 241: / Line 241: @@
 <pre>
-import matplotlib.pyplot as plt
+>>> import matplotlib.pyplot as plt
-from pylab import meshgrid
+>>> from pylab import meshgrid
-X, Y = meshgrid(x,y)
+>>> X, Y = meshgrid(x,y)
-plt.figure()
+>>> plt.figure()
-plt.contour(X,Y,Z)
+>>> plt.contour(X,Y,Z)
-plt.show()
+>>> plt.show()
 </pre>

@@ Line 216: / Line 216: @@
 The use of '''resize''' in the last example illustrates a useful '''replicating feature'''.
-= Interest =
+=Visualising the Contents of an Array=
 <pre>

@@ Line 200: / Line 200: @@
 >>> print a
 </pre>
 = Interest =

← Older revision		Revision as of 16:01, 22 July 2009
Line 190:		Line 190:

	'''Note, however, that most operations on numpy arrays are done element-wise''', which may be '''different to a linear algera operation that you were expecting.''' We will return to linear algebra operations in a later section.		'''Note, however, that most operations on numpy arrays are done element-wise''', which may be '''different to a linear algera operation that you were expecting.''' We will return to linear algebra operations in a later section.
		+
		+	As with the list example, it can be useful to read or change the value of an element (or sub array) indidually. Let's turn the array back to it's rank-2 form and try it out:
		+
		+	<pre>
		+	>>> a.shape = (3,3)
		+	>>> a[1,1] = 777.0
		+	>>> print a
		+	>>> a[1:,1:] = [[777.0, 777.0],[777.0, 777.0]]
		+	>>> print a
		+	</pre>

	= Interest =		= Interest =

@@ Line 191: / Line 191: @@
 '''Note, however, that most operations on numpy arrays are done element-wise''', which may be '''different to a linear algera operation that you were expecting.'''  We will return to linear algebra operations in a later section.
-=Arrays=
+= Interest =
 <pre>

@@ Line 112: / Line 112: @@
 <pre>
 >>> len(shopping)
 </pre>