Show More
| This diff has been collapsed as it changes many lines, (751 lines changed) Show them Hide them | |||
| @@ -1,379 +1,378 b'' | |||
|
|
1 | 1 | { |
|
|
2 |
|
|
|
|
3 |
|
|
|
|
2 | "metadata": { | |
|
|
3 | "name": "display_protocol" | |
|
|
4 | }, | |
|
|
5 | "nbformat": 2, | |
|
|
6 | "worksheets": [ | |
|
|
7 | { | |
|
|
8 | "cells": [ | |
|
|
9 | { | |
|
|
10 | "cell_type": "markdown", | |
|
|
11 | "source": [ | |
|
|
12 | "# Using the IPython display protocol for your own objects", | |
|
|
13 | "", | |
|
|
14 | "IPython extends the idea of the ``__repr__`` method in Python to support multiple representations for a given", | |
|
|
15 | "object, which clients can use to display the object according to their capabilities. An object can return multiple", | |
|
|
16 | "representations of itself by implementing special methods, and you can also define at runtime custom display ", | |
|
|
17 | "functions for existing objects whose methods you can't or won't modify. In this notebook, we show how both approaches work.", | |
|
|
18 | "", | |
|
|
19 | "<br/>", | |
|
|
20 | "**Note:** this notebook has had all output cells stripped out so we can include it in the IPython documentation with ", | |
|
|
21 | "a minimal file size. You'll need to manually execute the cells to see the output (you can run all of them with the ", | |
|
|
22 | "\"Run All\" button, or execute each individually). You must start this notebook with", | |
|
|
23 | "<pre>", | |
|
|
24 | "ipython notebook --pylab inline", | |
|
|
25 | "</pre>", | |
|
|
26 | "", | |
|
|
27 | "to ensure pylab support is available for plots.", | |
|
|
28 | "", | |
|
|
29 | "## Custom-built classes with dedicated ``_repr_*_`` methods", | |
|
|
30 | "", | |
|
|
31 | "In our first example, we illustrate how objects can expose directly to IPython special representations of", | |
|
|
32 | "themselves, by providing methods such as ``_repr_svg_``, ``_repr_png_``, ``_repr_latex_``, etc. For a full", | |
|
|
33 | "list of the special ``_repr_*_`` methods supported, see the code in ``IPython.core.displaypub``.", | |
|
|
34 | "", | |
|
|
35 | "As an illustration, we build a class that holds data generated by sampling a Gaussian distribution with given mean ", | |
|
|
36 | "and variance. The class can display itself in a variety of ways: as a LaTeX expression or as an image in PNG or SVG ", | |
|
|
37 | "format. Each frontend can then decide which representation it can handle.", | |
|
|
38 | "Further, we illustrate how to expose directly to the user the ability to directly access the various alternate ", | |
|
|
39 | "representations (since by default displaying the object itself will only show one, and which is shown will depend on the ", | |
|
|
40 | "required representations that even cache necessary data in cases where it may be expensive to compute.", | |
|
|
41 | "", | |
|
|
42 | "The next cell defines the Gaussian class:" | |
|
|
43 | ] | |
|
|
4 | 44 | }, |
|
|
5 | "nbformat": 2, | |
|
|
6 | "worksheets": [ | |
|
|
7 | { | |
|
|
8 | "cells": [ | |
|
|
9 | { | |
|
|
10 | "cell_type": "markdown", | |
|
|
11 | "source": [ | |
|
|
12 | "# Using the IPython display protocol for your own objects", | |
|
|
13 | "", | |
|
|
14 | "IPython extends the idea of the ``__repr__`` method in Python to support multiple representations for a given", | |
|
|
15 | "object, which clients can use to display the object according to their capabilities. An object can return multiple", | |
|
|
16 | "representations of itself by implementing special methods, and you can also define at runtime custom display ", | |
|
|
17 | "functions for existing objects whose methods you can't or won't modify. In this notebook, we show how both approaches work.", | |
|
|
18 | "", | |
|
|
19 | "<br/>", | |
|
|
20 | "**Note:** this notebook has had all output cells stripped out so we can include it in the IPython documentation with ", | |
|
|
21 | "a minimal file size. You'll need to manually execute the cells to see the output (you can run all of them with the ", | |
|
|
22 | "\"Run All\" button, or execute each individually). You must start this notebook with", | |
|
|
23 |
|
|
|
|
24 | "ipython notebook --pylab inline", | |
|
|
25 | "</pre>", | |
|
|
26 | "", | |
|
|
27 | "to ensure pylab support is available for plots.", | |
|
|
28 | "", | |
|
|
29 | "## Custom-built classes with dedicated ``_repr_*_`` methods", | |
|
|
30 | "", | |
|
|
31 | "In our first example, we illustrate how objects can expose directly to IPython special representations of", | |
|
|
32 | "themselves, by providing methods such as ``_repr_svg_``, ``_repr_png_``, ``_repr_latex_``, etc. For a full", | |
|
|
33 | "list of the special ``_repr_*_`` methods supported, see the code in ``IPython.core.displaypub``.", | |
|
|
34 | "", | |
|
|
35 | "As an illustration, we build a class that holds data generated by sampling a Gaussian distribution with given mean ", | |
|
|
36 | "and variance. The class can display itself in a variety of ways: as a LaTeX expression or as an image in PNG or SVG ", | |
|
|
37 | "format. Each frontend can then decide which representation it can handle.", | |
|
|
38 | "Further, we illustrate how to expose directly to the user the ability to directly access the various alternate ", | |
|
|
39 | "representations (since by default displaying the object itself will only show one, and which is shown will depend on the ", | |
|
|
40 | "required representations that even cache necessary data in cases where it may be expensive to compute.", | |
|
|
41 | "", | |
|
|
42 | "The next cell defines the Gaussian class:" | |
|
|
43 | ] | |
|
|
44 | }, | |
|
|
45 | { | |
|
|
46 | "cell_type": "code", | |
|
|
47 | "collapsed": true, | |
|
|
48 | "input": [ | |
|
|
49 | "from IPython.lib.pylabtools import print_figure", | |
|
|
50 | "from IPython.core.display import Image, SVG, Math", | |
|
|
51 | "", | |
|
|
52 | "class Gaussian(object):", | |
|
|
53 | " \"\"\"A simple object holding data sampled from a Gaussian distribution.", | |
|
|
54 | " \"\"\"", | |
|
|
55 | " def __init__(self, mean=0, std=1, size=1000):", | |
|
|
56 | " self.data = np.random.normal(mean, std, size)", | |
|
|
57 | " self.mean = mean", | |
|
|
58 | " self.std = std", | |
|
|
59 | " self.size = size", | |
|
|
60 | " # For caching plots that may be expensive to compute", | |
|
|
61 | " self._png_data = None", | |
|
|
62 | " self._svg_data = None", | |
|
|
63 | " ", | |
|
|
64 | " def _figure_data(self, format):", | |
|
|
65 | " fig, ax = plt.subplots()", | |
|
|
66 | " ax.plot(self.data, 'o')", | |
|
|
67 | " ax.set_title(self._repr_latex_())", | |
|
|
68 | " data = print_figure(fig, format)", | |
|
|
69 | " # We MUST close the figure, otherwise IPython's display machinery", | |
|
|
70 | " # will pick it up and send it as output, resulting in a double display", | |
|
|
71 | " plt.close(fig)", | |
|
|
72 | " return data", | |
|
|
73 | " ", | |
|
|
74 | " # Here we define the special repr methods that provide the IPython display protocol", | |
|
|
75 | " # Note that for the two figures, we cache the figure data once computed.", | |
|
|
76 | " ", | |
|
|
77 | " def _repr_png_(self):", | |
|
|
78 | " if self._png_data is None:", | |
|
|
79 | " self._png_data = self._figure_data('png')", | |
|
|
80 | " return self._png_data", | |
|
|
81 | "", | |
|
|
82 | "", | |
|
|
83 | " def _repr_svg_(self):", | |
|
|
84 | " if self._svg_data is None:", | |
|
|
85 | " self._svg_data = self._figure_data('svg')", | |
|
|
86 | " return self._svg_data", | |
|
|
87 | " ", | |
|
|
88 | " def _repr_latex_(self):", | |
|
|
89 | " return r'$\\mathcal{N}(\\mu=%.2g, \\sigma=%.2g),\\ N=%d$' % (self.mean,", | |
|
|
90 | " self.std, self.size)", | |
|
|
91 | " ", | |
|
|
92 | " # We expose as properties some of the above reprs, so that the user can see them", | |
|
|
93 | " # directly (since otherwise the client dictates which one it shows by default)", | |
|
|
94 | " @property", | |
|
|
95 | " def png(self):", | |
|
|
96 | " return Image(self._repr_png_(), embed=True)", | |
|
|
97 | " ", | |
|
|
98 | " @property", | |
|
|
99 | " def svg(self):", | |
|
|
100 | " return SVG(self._repr_svg_())", | |
|
|
101 | " ", | |
|
|
102 | " @property", | |
|
|
103 | " def latex(self):", | |
|
|
104 | " return Math(self._repr_svg_())", | |
|
|
105 | " ", | |
|
|
106 | " # An example of using a property to display rich information, in this case", | |
|
|
107 | " # the histogram of the distribution. We've hardcoded the format to be png", | |
|
|
108 | " # in this case, but in production code it would be trivial to make it an option", | |
|
|
109 | " @property", | |
|
|
110 | " def hist(self):", | |
|
|
111 | " fig, ax = plt.subplots()", | |
|
|
112 | " ax.hist(self.data, bins=100)", | |
|
|
113 | " ax.set_title(self._repr_latex_())", | |
|
|
114 | " data = print_figure(fig, 'png')", | |
|
|
115 | " plt.close(fig)", | |
|
|
116 | " return Image(data, embed=True)" | |
|
|
117 | ], | |
|
|
118 | "language": "python", | |
|
|
119 | "outputs": [], | |
|
|
120 | "prompt_number": 1 | |
|
|
121 | }, | |
|
|
122 | { | |
|
|
123 | "cell_type": "markdown", | |
|
|
124 | "source": [ | |
|
|
125 | "Now, we create an instance of the Gaussian distribution, whose default representation will be its LaTeX form:" | |
|
|
126 | ] | |
|
|
127 | }, | |
|
|
128 | { | |
|
|
129 | "cell_type": "code", | |
|
|
130 | "collapsed": true, | |
|
|
131 | "input": [ | |
|
|
132 | "x = Gaussian()", | |
|
|
133 | "x" | |
|
|
134 | ], | |
|
|
135 | "language": "python", | |
|
|
136 | "outputs": [], | |
|
|
137 | "prompt_number": 2 | |
|
|
138 | }, | |
|
|
139 | { | |
|
|
140 | "cell_type": "markdown", | |
|
|
141 | "source": [ | |
|
|
142 | "We can view the data in png or svg formats:" | |
|
|
143 | ] | |
|
|
144 | }, | |
|
|
145 | { | |
|
|
146 | "cell_type": "code", | |
|
|
147 | "collapsed": true, | |
|
|
148 | "input": [ | |
|
|
149 | "x.png" | |
|
|
150 | ], | |
|
|
151 | "language": "python", | |
|
|
152 | "outputs": [], | |
|
|
153 | "prompt_number": 3 | |
|
|
154 | }, | |
|
|
155 | { | |
|
|
156 | "cell_type": "code", | |
|
|
157 | "collapsed": true, | |
|
|
158 | "input": [ | |
|
|
159 | "x.svg" | |
|
|
160 | ], | |
|
|
161 | "language": "python", | |
|
|
162 | "outputs": [], | |
|
|
163 | "prompt_number": 4 | |
|
|
164 | }, | |
|
|
165 | { | |
|
|
166 | "cell_type": "markdown", | |
|
|
167 | "source": [ | |
|
|
168 | "Since IPython only displays by default as an ``Out[]`` cell the result of the last computation, we can use the", | |
|
|
169 | "``display()`` function to show more than one representation in a single cell:" | |
|
|
170 | ] | |
|
|
171 | }, | |
|
|
172 | { | |
|
|
173 | "cell_type": "code", | |
|
|
174 | "collapsed": true, | |
|
|
175 | "input": [ | |
|
|
176 | "display(x.png)", | |
|
|
177 | "display(x.svg)" | |
|
|
178 | ], | |
|
|
179 | "language": "python", | |
|
|
180 | "outputs": [], | |
|
|
181 | "prompt_number": 5 | |
|
|
182 | }, | |
|
|
183 | { | |
|
|
184 | "cell_type": "markdown", | |
|
|
185 | "source": [ | |
|
|
186 | "Now let's create a new Gaussian with different parameters" | |
|
|
187 | ] | |
|
|
188 | }, | |
|
|
189 | { | |
|
|
190 | "cell_type": "code", | |
|
|
191 | "collapsed": true, | |
|
|
192 | "input": [ | |
|
|
193 | "x2 = Gaussian(0.5, 0.2, 2000)", | |
|
|
194 | "x2" | |
|
|
195 | ], | |
|
|
196 | "language": "python", | |
|
|
197 | "outputs": [], | |
|
|
198 | "prompt_number": 6 | |
|
|
199 | }, | |
|
|
200 | { | |
|
|
201 | "cell_type": "markdown", | |
|
|
202 | "source": [ | |
|
|
203 | "We can easily compare them by displaying their histograms" | |
|
|
204 | ] | |
|
|
205 | }, | |
|
|
206 | { | |
|
|
207 | "cell_type": "code", | |
|
|
208 | "collapsed": true, | |
|
|
209 | "input": [ | |
|
|
210 | "display(x.hist)", | |
|
|
211 | "display(x2.hist)" | |
|
|
212 | ], | |
|
|
213 | "language": "python", | |
|
|
214 | "outputs": [], | |
|
|
215 | "prompt_number": 7 | |
|
|
216 |
|
|
|
|
217 | { | |
|
|
218 | "cell_type": "markdown", | |
|
|
219 | "source": [ | |
|
|
220 | "## Adding IPython display support to existing objects", | |
|
|
221 | "", | |
|
|
222 | "When you are directly writing your own classes, you can adapt them for display in IPython by ", | |
|
|
223 | "following the above example. But in practice, we often need to work with existing code we", | |
|
|
224 | "can't modify. ", | |
|
|
225 | "", | |
|
|
226 | "We now illustrate how to add these kinds of extended display capabilities to existing objects.", | |
|
|
227 | "We will use the numpy polynomials and change their default representation to be a formatted", | |
|
|
228 | "LaTeX expression.", | |
|
|
229 | "", | |
|
|
230 | "First, consider how a numpy polynomial object renders by default:" | |
|
|
231 | ] | |
|
|
232 | }, | |
|
|
233 | { | |
|
|
234 | "cell_type": "code", | |
|
|
235 | "collapsed": true, | |
|
|
236 | "input": [ | |
|
|
237 | "p = np.polynomial.Polynomial([1,2,3], [-10, 10])", | |
|
|
238 | "p" | |
|
|
239 | ], | |
|
|
240 | "language": "python", | |
|
|
241 | "outputs": [], | |
|
|
242 | "prompt_number": 8 | |
|
|
243 | }, | |
|
|
244 | { | |
|
|
245 | "cell_type": "markdown", | |
|
|
246 | "source": [ | |
|
|
247 | "Next, we define a function that pretty-prints a polynomial as a LaTeX string:" | |
|
|
248 | ] | |
|
|
249 | }, | |
|
|
250 | { | |
|
|
251 | "cell_type": "code", | |
|
|
252 | "collapsed": true, | |
|
|
253 | "input": [ | |
|
|
254 | "def poly2latex(p):", | |
|
|
255 | " terms = ['%.2g' % p.coef[0]]", | |
|
|
256 | " if len(p) > 1:", | |
|
|
257 | " term = 'x'", | |
|
|
258 | " c = p.coef[1]", | |
|
|
259 | " if c!=1:", | |
|
|
260 | " term = ('%.2g ' % c) + term", | |
|
|
261 | " terms.append(term)", | |
|
|
262 | " if len(p) > 2:", | |
|
|
263 | " for i in range(2, len(p)):", | |
|
|
264 | " term = 'x^%d' % i", | |
|
|
265 | " c = p.coef[i]", | |
|
|
266 | " if c!=1:", | |
|
|
267 | " term = ('%.2g ' % c) + term", | |
|
|
268 | " terms.append(term)", | |
|
|
269 | " px = '$P(x)=%s$' % '+'.join(terms)", | |
|
|
270 | " dom = r', domain: $[%.2g,\\ %.2g]$' % tuple(p.domain)", | |
|
|
271 | " win = r', window: $[%.2g,\\ %.2g]$' % tuple(p.window)", | |
|
|
272 | " return px+dom+win" | |
|
|
273 | ], | |
|
|
274 | "language": "python", | |
|
|
275 | "outputs": [], | |
|
|
276 | "prompt_number": 9 | |
|
|
277 | }, | |
|
|
278 | { | |
|
|
279 | "cell_type": "markdown", | |
|
|
280 | "source": [ | |
|
|
281 | "This produces, on our polynomial ``p``, the following:" | |
|
|
282 | ] | |
|
|
283 | }, | |
|
|
284 | { | |
|
|
285 | "cell_type": "code", | |
|
|
286 | "collapsed": true, | |
|
|
287 | "input": [ | |
|
|
288 |
|
|
|
|
289 | ], | |
|
|
290 |
|
|
|
|
291 |
|
|
|
|
292 |
|
|
|
|
293 | }, | |
|
|
294 | { | |
|
|
295 |
|
|
|
|
296 |
|
|
|
|
297 | "Note that this did *not* produce a formated LaTeX object, because it is simply a string ", | |
|
|
298 | "with LaTeX code. In order for this to be interpreted as a mathematical expression, it", | |
|
|
299 | "must be properly wrapped into a Math object:" | |
|
|
300 | ] | |
|
|
301 | }, | |
|
|
302 | { | |
|
|
303 | "cell_type": "code", | |
|
|
304 | "collapsed": true, | |
|
|
305 | "input": [ | |
|
|
306 | "from IPython.core.display import Math", | |
|
|
307 | "Math(poly2latex(p))" | |
|
|
308 | ], | |
|
|
309 | "language": "python", | |
|
|
310 | "outputs": [], | |
|
|
311 | "prompt_number": 11 | |
|
|
312 | }, | |
|
|
313 | { | |
|
|
314 | "cell_type": "markdown", | |
|
|
315 | "source": [ | |
|
|
316 | "But we can configure IPython to do this automatically for us as follows. We hook into the", | |
|
|
317 | "IPython display system and instruct it to use ``poly2latex`` for the latex mimetype, when", | |
|
|
318 | "encountering objects of the ``Polynomial`` type defined in the", | |
|
|
319 | "``numpy.polynomial.polynomial`` module:" | |
|
|
320 | ] | |
|
|
321 | }, | |
|
|
322 | { | |
|
|
323 | "cell_type": "code", | |
|
|
324 | "collapsed": true, | |
|
|
325 | "input": [ | |
|
|
326 | "ip = get_ipython()", | |
|
|
327 | "latex_formatter = ip.display_formatter.formatters['text/latex']", | |
|
|
328 | "latex_formatter.for_type_by_name('numpy.polynomial.polynomial',", | |
|
|
329 | " 'Polynomial', poly2latex)" | |
|
|
330 | ], | |
|
|
331 |
|
|
|
|
332 |
|
|
|
|
333 |
|
|
|
|
334 | }, | |
|
|
335 | { | |
|
|
336 | "cell_type": "markdown", | |
|
|
337 | "source": [ | |
|
|
338 | "For more examples on how to use the above system, and how to bundle similar print functions", | |
|
|
339 | "into a convenient IPython extension, see the ``IPython/extensions/sympyprinting.py`` file. ", | |
|
|
340 | "The machinery that defines the display system is in the ``display.py`` and ``displaypub.py`` ", | |
|
|
341 | "files in ``IPython/core``.", | |
|
|
342 | "", | |
|
|
343 | "Once our special printer has been loaded, all polynomials will be represented by their ", | |
|
|
344 | "mathematical form instead:" | |
|
|
345 | ] | |
|
|
346 | }, | |
|
|
347 | { | |
|
|
348 | "cell_type": "code", | |
|
|
349 | "collapsed": true, | |
|
|
350 | "input": [ | |
|
|
351 | "p" | |
|
|
352 | ], | |
|
|
353 | "language": "python", | |
|
|
354 | "outputs": [], | |
|
|
355 | "prompt_number": 13 | |
|
|
356 | }, | |
|
|
357 | { | |
|
|
358 | "cell_type": "code", | |
|
|
359 | "collapsed": true, | |
|
|
360 | "input": [ | |
|
|
361 | "p2 = np.polynomial.Polynomial([-20, 71, -15, 1])", | |
|
|
362 | "p2" | |
|
|
363 | ], | |
|
|
364 | "language": "python", | |
|
|
365 | "outputs": [], | |
|
|
366 | "prompt_number": 14 | |
|
|
367 | }, | |
|
|
368 | { | |
|
|
369 | "cell_type": "code", | |
|
|
370 | "collapsed": true, | |
|
|
371 | "input": [], | |
|
|
372 | "language": "python", | |
|
|
373 | "outputs": [], | |
|
|
374 | "prompt_number": 14 | |
|
|
375 | } | |
|
|
376 | ] | |
|
|
377 | } | |
|
|
378 | ] | |
|
|
45 | { | |
|
|
46 | "cell_type": "code", | |
|
|
47 | "collapsed": true, | |
|
|
48 | "input": [ | |
|
|
49 | "from IPython.lib.pylabtools import print_figure", | |
|
|
50 | "from IPython.core.display import Image, SVG, Math", | |
|
|
51 | "", | |
|
|
52 | "class Gaussian(object):", | |
|
|
53 | " \"\"\"A simple object holding data sampled from a Gaussian distribution.", | |
|
|
54 | " \"\"\"", | |
|
|
55 | " def __init__(self, mean=0, std=1, size=1000):", | |
|
|
56 | " self.data = np.random.normal(mean, std, size)", | |
|
|
57 | " self.mean = mean", | |
|
|
58 | " self.std = std", | |
|
|
59 | " self.size = size", | |
|
|
60 | " # For caching plots that may be expensive to compute", | |
|
|
61 | " self._png_data = None", | |
|
|
62 | " self._svg_data = None", | |
|
|
63 | " ", | |
|
|
64 | " def _figure_data(self, format):", | |
|
|
65 | " fig, ax = plt.subplots()", | |
|
|
66 | " ax.plot(self.data, 'o')", | |
|
|
67 | " ax.set_title(self._repr_latex_())", | |
|
|
68 | " data = print_figure(fig, format)", | |
|
|
69 | " # We MUST close the figure, otherwise IPython's display machinery", | |
|
|
70 | " # will pick it up and send it as output, resulting in a double display", | |
|
|
71 | " plt.close(fig)", | |
|
|
72 | " return data", | |
|
|
73 | " ", | |
|
|
74 | " # Here we define the special repr methods that provide the IPython display protocol", | |
|
|
75 | " # Note that for the two figures, we cache the figure data once computed.", | |
|
|
76 | " ", | |
|
|
77 | " def _repr_png_(self):", | |
|
|
78 | " if self._png_data is None:", | |
|
|
79 | " self._png_data = self._figure_data('png')", | |
|
|
80 | " return self._png_data", | |
|
|
81 | "", | |
|
|
82 | "", | |
|
|
83 | " def _repr_svg_(self):", | |
|
|
84 | " if self._svg_data is None:", | |
|
|
85 | " self._svg_data = self._figure_data('svg')", | |
|
|
86 | " return self._svg_data", | |
|
|
87 | " ", | |
|
|
88 | " def _repr_latex_(self):", | |
|
|
89 | " return r'$\\mathcal{N}(\\mu=%.2g, \\sigma=%.2g),\\ N=%d$' % (self.mean,", | |
|
|
90 | " self.std, self.size)", | |
|
|
91 | " ", | |
|
|
92 | " # We expose as properties some of the above reprs, so that the user can see them", | |
|
|
93 | " # directly (since otherwise the client dictates which one it shows by default)", | |
|
|
94 | " @property", | |
|
|
95 | " def png(self):", | |
|
|
96 | " return Image(self._repr_png_(), embed=True)", | |
|
|
97 | " ", | |
|
|
98 | " @property", | |
|
|
99 | " def svg(self):", | |
|
|
100 | " return SVG(self._repr_svg_())", | |
|
|
101 | " ", | |
|
|
102 | " @property", | |
|
|
103 | " def latex(self):", | |
|
|
104 | " return Math(self._repr_svg_())", | |
|
|
105 | " ", | |
|
|
106 | " # An example of using a property to display rich information, in this case", | |
|
|
107 | " # the histogram of the distribution. We've hardcoded the format to be png", | |
|
|
108 | " # in this case, but in production code it would be trivial to make it an option", | |
|
|
109 | " @property", | |
|
|
110 | " def hist(self):", | |
|
|
111 | " fig, ax = plt.subplots()", | |
|
|
112 | " ax.hist(self.data, bins=100)", | |
|
|
113 | " ax.set_title(self._repr_latex_())", | |
|
|
114 | " data = print_figure(fig, 'png')", | |
|
|
115 | " plt.close(fig)", | |
|
|
116 | " return Image(data, embed=True)" | |
|
|
117 | ], | |
|
|
118 | "language": "python", | |
|
|
119 | "outputs": [], | |
|
|
120 | "prompt_number": 1 | |
|
|
121 | }, | |
|
|
122 | { | |
|
|
123 | "cell_type": "markdown", | |
|
|
124 | "source": [ | |
|
|
125 | "Now, we create an instance of the Gaussian distribution, whose default representation will be its LaTeX form:" | |
|
|
126 | ] | |
|
|
127 | }, | |
|
|
128 | { | |
|
|
129 | "cell_type": "code", | |
|
|
130 | "collapsed": false, | |
|
|
131 | "input": [ | |
|
|
132 | "x = Gaussian()", | |
|
|
133 | "x" | |
|
|
134 | ], | |
|
|
135 | "language": "python", | |
|
|
136 | "outputs": [], | |
|
|
137 | "prompt_number": 2 | |
|
|
138 | }, | |
|
|
139 | { | |
|
|
140 | "cell_type": "markdown", | |
|
|
141 | "source": [ | |
|
|
142 | "We can view the data in png or svg formats:" | |
|
|
143 | ] | |
|
|
144 | }, | |
|
|
145 | { | |
|
|
146 | "cell_type": "code", | |
|
|
147 | "collapsed": false, | |
|
|
148 | "input": [ | |
|
|
149 | "x.png" | |
|
|
150 | ], | |
|
|
151 | "language": "python", | |
|
|
152 | "outputs": [], | |
|
|
153 | "prompt_number": 3 | |
|
|
154 | }, | |
|
|
155 | { | |
|
|
156 | "cell_type": "code", | |
|
|
157 | "collapsed": false, | |
|
|
158 | "input": [ | |
|
|
159 | "x.svg" | |
|
|
160 | ], | |
|
|
161 | "language": "python", | |
|
|
162 | "outputs": [], | |
|
|
163 | "prompt_number": 4 | |
|
|
164 | }, | |
|
|
165 | { | |
|
|
166 | "cell_type": "markdown", | |
|
|
167 | "source": [ | |
|
|
168 | "Since IPython only displays by default as an ``Out[]`` cell the result of the last computation, we can use the", | |
|
|
169 | "``display()`` function to show more than one representation in a single cell:" | |
|
|
170 | ] | |
|
|
171 | }, | |
|
|
172 | { | |
|
|
173 | "cell_type": "code", | |
|
|
174 | "collapsed": false, | |
|
|
175 | "input": [ | |
|
|
176 | "display(x.png)", | |
|
|
177 | "display(x.svg)" | |
|
|
178 | ], | |
|
|
179 | "language": "python", | |
|
|
180 | "outputs": [], | |
|
|
181 | "prompt_number": 5 | |
|
|
182 | }, | |
|
|
183 | { | |
|
|
184 | "cell_type": "markdown", | |
|
|
185 | "source": [ | |
|
|
186 | "Now let's create a new Gaussian with different parameters" | |
|
|
187 | ] | |
|
|
188 | }, | |
|
|
189 | { | |
|
|
190 | "cell_type": "code", | |
|
|
191 | "collapsed": false, | |
|
|
192 | "input": [ | |
|
|
193 | "x2 = Gaussian(0.5, 0.2, 2000)", | |
|
|
194 | "x2" | |
|
|
195 | ], | |
|
|
196 | "language": "python", | |
|
|
197 | "outputs": [], | |
|
|
198 | "prompt_number": 6 | |
|
|
199 | }, | |
|
|
200 | { | |
|
|
201 | "cell_type": "markdown", | |
|
|
202 | "source": [ | |
|
|
203 | "We can easily compare them by displaying their histograms" | |
|
|
204 | ] | |
|
|
205 | }, | |
|
|
206 | { | |
|
|
207 | "cell_type": "code", | |
|
|
208 | "collapsed": false, | |
|
|
209 | "input": [ | |
|
|
210 | "display(x.hist)", | |
|
|
211 | "display(x2.hist)" | |
|
|
212 | ], | |
|
|
213 | "language": "python", | |
|
|
214 | "outputs": [], | |
|
|
215 | "prompt_number": 7 | |
|
|
216 | }, | |
|
|
217 | { | |
|
|
218 | "cell_type": "markdown", | |
|
|
219 | "source": [ | |
|
|
220 | "## Adding IPython display support to existing objects", | |
|
|
221 | "", | |
|
|
222 | "When you are directly writing your own classes, you can adapt them for display in IPython by ", | |
|
|
223 | "following the above example. But in practice, we often need to work with existing code we", | |
|
|
224 | "can't modify. ", | |
|
|
225 | "", | |
|
|
226 | "We now illustrate how to add these kinds of extended display capabilities to existing objects.", | |
|
|
227 | "We will use the numpy polynomials and change their default representation to be a formatted", | |
|
|
228 | "LaTeX expression.", | |
|
|
229 | "", | |
|
|
230 | "First, consider how a numpy polynomial object renders by default:" | |
|
|
231 | ] | |
|
|
232 | }, | |
|
|
233 | { | |
|
|
234 | "cell_type": "code", | |
|
|
235 | "collapsed": false, | |
|
|
236 | "input": [ | |
|
|
237 | "p = np.polynomial.Polynomial([1,2,3], [-10, 10])", | |
|
|
238 | "p" | |
|
|
239 | ], | |
|
|
240 | "language": "python", | |
|
|
241 | "outputs": [], | |
|
|
242 | "prompt_number": 8 | |
|
|
243 | }, | |
|
|
244 | { | |
|
|
245 | "cell_type": "markdown", | |
|
|
246 | "source": [ | |
|
|
247 | "Next, we define a function that pretty-prints a polynomial as a LaTeX string:" | |
|
|
248 | ] | |
|
|
249 | }, | |
|
|
250 | { | |
|
|
251 | "cell_type": "code", | |
|
|
252 | "collapsed": true, | |
|
|
253 | "input": [ | |
|
|
254 | "def poly2latex(p):", | |
|
|
255 | " terms = ['%.2g' % p.coef[0]]", | |
|
|
256 | " if len(p) > 1:", | |
|
|
257 | " term = 'x'", | |
|
|
258 | " c = p.coef[1]", | |
|
|
259 | " if c!=1:", | |
|
|
260 | " term = ('%.2g ' % c) + term", | |
|
|
261 | " terms.append(term)", | |
|
|
262 | " if len(p) > 2:", | |
|
|
263 | " for i in range(2, len(p)):", | |
|
|
264 | " term = 'x^%d' % i", | |
|
|
265 | " c = p.coef[i]", | |
|
|
266 | " if c!=1:", | |
|
|
267 | " term = ('%.2g ' % c) + term", | |
|
|
268 | " terms.append(term)", | |
|
|
269 | " px = '$P(x)=%s$' % '+'.join(terms)", | |
|
|
270 | " dom = r', domain: $[%.2g,\\ %.2g]$' % tuple(p.domain)", | |
|
|
271 | " return px+dom" | |
|
|
272 | ], | |
|
|
273 | "language": "python", | |
|
|
274 | "outputs": [], | |
|
|
275 | "prompt_number": 11 | |
|
|
276 | }, | |
|
|
277 | { | |
|
|
278 | "cell_type": "markdown", | |
|
|
279 | "source": [ | |
|
|
280 | "This produces, on our polynomial ``p``, the following:" | |
|
|
281 | ] | |
|
|
282 | }, | |
|
|
283 | { | |
|
|
284 | "cell_type": "code", | |
|
|
285 | "collapsed": false, | |
|
|
286 | "input": [ | |
|
|
287 | "poly2latex(p)" | |
|
|
288 | ], | |
|
|
289 | "language": "python", | |
|
|
290 | "outputs": [], | |
|
|
291 | "prompt_number": 12 | |
|
|
292 | }, | |
|
|
293 | { | |
|
|
294 | "cell_type": "markdown", | |
|
|
295 | "source": [ | |
|
|
296 | "Note that this did *not* produce a formated LaTeX object, because it is simply a string ", | |
|
|
297 | "with LaTeX code. In order for this to be interpreted as a mathematical expression, it", | |
|
|
298 | "must be properly wrapped into a Math object:" | |
|
|
299 | ] | |
|
|
300 | }, | |
|
|
301 | { | |
|
|
302 | "cell_type": "code", | |
|
|
303 | "collapsed": false, | |
|
|
304 | "input": [ | |
|
|
305 | "from IPython.core.display import Math", | |
|
|
306 | "Math(poly2latex(p))" | |
|
|
307 | ], | |
|
|
308 | "language": "python", | |
|
|
309 | "outputs": [], | |
|
|
310 | "prompt_number": 13 | |
|
|
311 | }, | |
|
|
312 | { | |
|
|
313 | "cell_type": "markdown", | |
|
|
314 | "source": [ | |
|
|
315 | "But we can configure IPython to do this automatically for us as follows. We hook into the", | |
|
|
316 | "IPython display system and instruct it to use ``poly2latex`` for the latex mimetype, when", | |
|
|
317 | "encountering objects of the ``Polynomial`` type defined in the", | |
|
|
318 | "``numpy.polynomial.polynomial`` module:" | |
|
|
319 | ] | |
|
|
320 | }, | |
|
|
321 | { | |
|
|
322 | "cell_type": "code", | |
|
|
323 | "collapsed": true, | |
|
|
324 | "input": [ | |
|
|
325 | "ip = get_ipython()", | |
|
|
326 | "latex_formatter = ip.display_formatter.formatters['text/latex']", | |
|
|
327 | "latex_formatter.for_type_by_name('numpy.polynomial.polynomial',", | |
|
|
328 | " 'Polynomial', poly2latex)" | |
|
|
329 | ], | |
|
|
330 | "language": "python", | |
|
|
331 | "outputs": [], | |
|
|
332 | "prompt_number": 14 | |
|
|
333 | }, | |
|
|
334 | { | |
|
|
335 | "cell_type": "markdown", | |
|
|
336 | "source": [ | |
|
|
337 | "For more examples on how to use the above system, and how to bundle similar print functions", | |
|
|
338 | "into a convenient IPython extension, see the ``IPython/extensions/sympyprinting.py`` file. ", | |
|
|
339 | "The machinery that defines the display system is in the ``display.py`` and ``displaypub.py`` ", | |
|
|
340 | "files in ``IPython/core``.", | |
|
|
341 | "", | |
|
|
342 | "Once our special printer has been loaded, all polynomials will be represented by their ", | |
|
|
343 | "mathematical form instead:" | |
|
|
344 | ] | |
|
|
345 | }, | |
|
|
346 | { | |
|
|
347 | "cell_type": "code", | |
|
|
348 | "collapsed": false, | |
|
|
349 | "input": [ | |
|
|
350 | "p" | |
|
|
351 | ], | |
|
|
352 | "language": "python", | |
|
|
353 | "outputs": [], | |
|
|
354 | "prompt_number": 15 | |
|
|
355 | }, | |
|
|
356 | { | |
|
|
357 | "cell_type": "code", | |
|
|
358 | "collapsed": false, | |
|
|
359 | "input": [ | |
|
|
360 | "p2 = np.polynomial.Polynomial([-20, 71, -15, 1])", | |
|
|
361 | "p2" | |
|
|
362 | ], | |
|
|
363 | "language": "python", | |
|
|
364 | "outputs": [], | |
|
|
365 | "prompt_number": 16 | |
|
|
366 | }, | |
|
|
367 | { | |
|
|
368 | "cell_type": "code", | |
|
|
369 | "collapsed": true, | |
|
|
370 | "input": [], | |
|
|
371 | "language": "python", | |
|
|
372 | "outputs": [], | |
|
|
373 | "prompt_number": 14 | |
|
|
374 | } | |
|
|
375 | ] | |
|
|
376 | } | |
|
|
377 | ] | |
|
|
379 | 378 | } No newline at end of file |
General Comments 0
You need to be logged in to leave comments.
Login now