prolly 0.0.1 → 0.0.2
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +4 -4
- data/README.markdown +286 -173
- data/lib/prolly/ps.rb +3 -3
- data/lib/prolly/ps/storage/rubylist.rb +1 -3
- metadata +1 -1
checksums.yaml
CHANGED
|
@@ -1,7 +1,7 @@
|
|
|
1
1
|
---
|
|
2
2
|
SHA1:
|
|
3
|
-
metadata.gz:
|
|
4
|
-
data.tar.gz:
|
|
3
|
+
metadata.gz: 2c026ea02e8ac2de33f7278ac2fe87e3d8864cf0
|
|
4
|
+
data.tar.gz: 7f9c5e4186921a36dcf74fe7662aa1b58551fd3e
|
|
5
5
|
SHA512:
|
|
6
|
-
metadata.gz:
|
|
7
|
-
data.tar.gz:
|
|
6
|
+
metadata.gz: 0bdfe8480833e3f61552ae656763f24acff8e8ca331e4c3d5e452286ed29f0b1e537067730fa1df043beec4e680c04c530f0151f0d025c282c837dceedf678f3
|
|
7
|
+
data.tar.gz: 6b5fcc2c664c0c18c123c0bc8c81bf109623293732696a886db3880212ee1f38898343f463b2e322180bb4f860f3b3d5d4cc959c179d2c63800dfd9e4bbf09a2
|
data/README.markdown
CHANGED
|
@@ -6,9 +6,10 @@ specifically for answering questions about probabilities of events based on the
|
|
|
6
6
|
samples you've seen before.
|
|
7
7
|
|
|
8
8
|
So instead of counting all the events yourself, you just express
|
|
9
|
-
probabilities much like how math books express it.
|
|
10
|
-
probabilities is useful for writing machine learning
|
|
11
|
-
of abstraction. The right level abstraction makes things
|
|
9
|
+
probabilities, entropies, and information gain much like how math books express it.
|
|
10
|
+
Being able to express probabilities is useful for writing machine learning
|
|
11
|
+
algorithms at a higher level of abstraction. The right level abstraction makes things
|
|
12
|
+
easier to build.
|
|
12
13
|
|
|
13
14
|
We can now making decisions in code not just based on the current data, like `if`
|
|
14
15
|
statements do, but we can make decisions based on the chance of prior data and
|
|
@@ -90,17 +91,99 @@ Ps.rv(color: :blue).given(size: :small).prob
|
|
|
90
91
|
```
|
|
91
92
|
And that will give you the probability of the random variable Color is :blue given that the Size was :small.
|
|
92
93
|
|
|
93
|
-
###
|
|
94
|
+
### Probabilities
|
|
95
|
+
|
|
96
|
+
What is the probability there is a blue marble?
|
|
97
|
+
```ruby
|
|
98
|
+
# P(C = blue)
|
|
99
|
+
Ps.rv(color: :blue).prob
|
|
100
|
+
```
|
|
101
|
+
|
|
102
|
+
What is the joint probability there is a blue marble that also has a rough texture?
|
|
103
|
+
```ruby
|
|
104
|
+
# P(C = blue, T = rough)
|
|
105
|
+
Ps.rv(color: :blue, texture: :rough).prob
|
|
106
|
+
```
|
|
107
|
+
|
|
108
|
+
What is the probability a marble is small or med sized?
|
|
109
|
+
```ruby
|
|
110
|
+
# P(S = small, med)
|
|
111
|
+
Ps.rv(size: [:small, :med]).prob
|
|
112
|
+
```
|
|
113
|
+
|
|
114
|
+
What is the probability of a blue marble given that the marble is small?
|
|
115
|
+
```ruby
|
|
116
|
+
# P(C = blue | S = small)
|
|
117
|
+
Ps.rv(color: :blue).given(size: :small).prob
|
|
118
|
+
```
|
|
119
|
+
|
|
120
|
+
What is the probability of a blue marble and rough texture given that the marble is small?
|
|
121
|
+
```ruby
|
|
122
|
+
# P(C = blue, T = rough | S = small)
|
|
123
|
+
Ps.rv(color: :blue, texture: :rough).given(size: :small).prob
|
|
124
|
+
```
|
|
125
|
+
|
|
126
|
+
### Probability density functions
|
|
127
|
+
|
|
128
|
+
Probability density for a random variable.
|
|
129
|
+
```ruby
|
|
130
|
+
Ps.rv(:color).pdf
|
|
131
|
+
```
|
|
132
|
+
|
|
133
|
+
Probability density for a conditional random variable.
|
|
134
|
+
```ruby
|
|
135
|
+
Ps.rv(:color).given(size: :small).pdf
|
|
136
|
+
```
|
|
137
|
+
|
|
138
|
+
### Entropy
|
|
139
|
+
|
|
140
|
+
Entropy of the RV color.
|
|
141
|
+
```ruby
|
|
142
|
+
# H(C)
|
|
143
|
+
Ps.rv(:color).entropy
|
|
144
|
+
```
|
|
145
|
+
|
|
146
|
+
Entropy of color given the marble is small
|
|
147
|
+
```ruby
|
|
148
|
+
# H(C | S = small)
|
|
149
|
+
Ps.rv(:color).given(size: :small).entropy
|
|
150
|
+
```
|
|
151
|
+
|
|
152
|
+
### Information Gain
|
|
153
|
+
|
|
154
|
+
Information gain of color and size.
|
|
155
|
+
```ruby
|
|
156
|
+
# IG(C | S)
|
|
157
|
+
Ps.rv(:color).given(:size).infogain
|
|
158
|
+
```
|
|
159
|
+
|
|
160
|
+
Information gain of color and size, when we already know texture and opacity.
|
|
161
|
+
```ruby
|
|
162
|
+
# IG(C | S, T=smooth, O=opaque)
|
|
163
|
+
Ps.rv(:color).given(:size, { texture: :smooth, opacity: :opaque }).infogain
|
|
164
|
+
```
|
|
165
|
+
|
|
166
|
+
### Counts
|
|
167
|
+
|
|
168
|
+
At the base of all the probabilities are counts of stuff.
|
|
169
|
+
```ruby
|
|
170
|
+
Ps.rv(color: :blue).count
|
|
171
|
+
```
|
|
172
|
+
|
|
173
|
+
```ruby
|
|
174
|
+
Ps.rv(:color).given(:size).count
|
|
175
|
+
```
|
|
176
|
+
## Full Reference
|
|
94
177
|
|
|
95
178
|
A random variable can be specified `Ps.rv(:color)` or unspecified `Ps.rv(color: :blue)`. So too can conditional random variables be specified or unspecified.
|
|
96
179
|
|
|
97
180
|
Prolly currently supports five operations.
|
|
98
181
|
|
|
99
|
-
- .prob · Calculates probability, a fractional number representing the belief you have that an event will occur; based on the amount of evidence you've seen for that event.
|
|
100
|
-
- .pdf · Calculates probability density function, a hash of all possible probabilities for the random variable.
|
|
101
|
-
- .entropy · Calculates entropy, a fractional number representing the spikiness or smoothness of a density function, which implies how much information is in the random variable.
|
|
102
|
-
- .infogain · Calculates information gain, a fractional number representing the amount of information (that is, reduction in uncertainty) that knowing either variable provides about the other.
|
|
103
|
-
- .count · Counts the number of events satisfying the conditions.
|
|
182
|
+
- .prob() · Calculates probability, a fractional number representing the belief you have that an event will occur; based on the amount of evidence you've seen for that event.
|
|
183
|
+
- .pdf() · Calculates probability density function, a hash of all possible probabilities for the random variable.
|
|
184
|
+
- .entropy() · Calculates entropy, a fractional number representing the spikiness or smoothness of a density function, which implies how much information is in the random variable.
|
|
185
|
+
- .infogain() · Calculates information gain, a fractional number representing the amount of information (that is, reduction in uncertainty) that knowing either variable provides about the other.
|
|
186
|
+
- .count() · Counts the number of events satisfying the conditions.
|
|
104
187
|
|
|
105
188
|
Each of the operations will only work with certain combinations of random variables. The possibilities are listed below, and Prolly will throw an exception if it's violated.
|
|
106
189
|
|
|
@@ -108,271 +191,301 @@ Legend:
|
|
|
108
191
|
- ✓ available for this operator
|
|
109
192
|
- Δ! available, but not yet implemented for this operator.
|
|
110
193
|
|
|
194
|
+
### The Probability Operator: .prob()
|
|
195
|
+
|
|
111
196
|
<table>
|
|
112
197
|
<tr>
|
|
113
|
-
<th
|
|
114
|
-
<th>
|
|
115
|
-
<th>.
|
|
116
|
-
<th>.
|
|
117
|
-
<th>.
|
|
118
|
-
<th>.
|
|
119
|
-
<th>.
|
|
198
|
+
<th></th>
|
|
199
|
+
<th>n/a</th>
|
|
200
|
+
<th>.given(:size)</th>
|
|
201
|
+
<th>.given(size: :small)</th>
|
|
202
|
+
<th>.given(size: :small, weight: :fat)</th>
|
|
203
|
+
<th>.given(:size, weight: :fat)</th>
|
|
204
|
+
<th>.given(:size, :weight)</th>
|
|
120
205
|
</tr>
|
|
121
206
|
<tr>
|
|
122
|
-
<th>
|
|
207
|
+
<th>rv(color: :blue)</th>
|
|
208
|
+
<th>✓</th>
|
|
209
|
+
<th>✓</th>
|
|
210
|
+
<th>✓</th>
|
|
211
|
+
<th>✓</th>
|
|
123
212
|
<th></th>
|
|
213
|
+
<th></th>
|
|
214
|
+
</tr>
|
|
215
|
+
<tr>
|
|
216
|
+
<th>rv(color: [:blue, :green])</th>
|
|
124
217
|
<th>✓</th>
|
|
125
218
|
<th></th>
|
|
126
219
|
<th></th>
|
|
127
220
|
<th></th>
|
|
128
|
-
<th
|
|
221
|
+
<th></th>
|
|
222
|
+
<th></th>
|
|
129
223
|
</tr>
|
|
130
224
|
<tr>
|
|
131
|
-
<th>
|
|
132
|
-
<th
|
|
225
|
+
<th>rv(color: :blue, texture: :rough)</th>
|
|
226
|
+
<th>✓</th>
|
|
133
227
|
<th>✓</th>
|
|
228
|
+
<th>✓</th>
|
|
229
|
+
<th>✓</th>
|
|
230
|
+
<th></th>
|
|
231
|
+
<th></th>
|
|
232
|
+
</tr>
|
|
233
|
+
<tr>
|
|
234
|
+
<th>rv(:color)</th>
|
|
235
|
+
<th></th>
|
|
236
|
+
<th></th>
|
|
237
|
+
<th></th>
|
|
134
238
|
<th></th>
|
|
135
239
|
<th></th>
|
|
136
240
|
<th></th>
|
|
137
|
-
<th>✓</th>
|
|
138
241
|
</tr>
|
|
139
242
|
<tr>
|
|
140
|
-
<th>
|
|
141
|
-
<th
|
|
142
|
-
<th
|
|
243
|
+
<th>rv(:color, :texture)</th>
|
|
244
|
+
<th></th>
|
|
245
|
+
<th></th>
|
|
246
|
+
<th></th>
|
|
143
247
|
<th></th>
|
|
144
248
|
<th></th>
|
|
145
249
|
<th></th>
|
|
146
|
-
<th>✓</th>
|
|
147
250
|
</tr>
|
|
251
|
+
</table>
|
|
252
|
+
|
|
253
|
+
### The Probability Density Function Operator: .pdf()
|
|
254
|
+
|
|
255
|
+
<table>
|
|
148
256
|
<tr>
|
|
149
|
-
<th
|
|
257
|
+
<th></th>
|
|
258
|
+
<th>n/a</th>
|
|
259
|
+
<th>.given(:size)</th>
|
|
260
|
+
<th>.given(size: :small)</th>
|
|
150
261
|
<th>.given(size: :small, weight: :fat)</th>
|
|
151
|
-
<th
|
|
262
|
+
<th>.given(:size, weight: :fat)</th>
|
|
263
|
+
<th>.given(:size, :weight)</th>
|
|
264
|
+
</tr>
|
|
265
|
+
<tr>
|
|
266
|
+
<th>rv(color: :blue)</th>
|
|
267
|
+
<th></th>
|
|
268
|
+
<th></th>
|
|
269
|
+
<th></th>
|
|
152
270
|
<th></th>
|
|
153
271
|
<th></th>
|
|
154
272
|
<th></th>
|
|
155
|
-
<th>✓</th>
|
|
156
273
|
</tr>
|
|
157
|
-
<tr>
|
|
158
|
-
<th>Ps.rv(color: [:blue, :green])</th>
|
|
159
|
-
<th></th>
|
|
160
|
-
<th>✓</th>
|
|
161
|
-
<th></th>
|
|
162
|
-
<th></th>
|
|
163
|
-
<th></th>
|
|
164
|
-
<th>✓</th>
|
|
165
|
-
</tr>
|
|
166
274
|
<tr>
|
|
167
|
-
<th>
|
|
275
|
+
<th>rv(color: [:blue, :green])</th>
|
|
276
|
+
<th></th>
|
|
277
|
+
<th></th>
|
|
168
278
|
<th></th>
|
|
169
|
-
<th>✓</th>
|
|
170
279
|
<th></th>
|
|
171
280
|
<th></th>
|
|
172
281
|
<th></th>
|
|
173
|
-
<th>✓</th>
|
|
174
282
|
</tr>
|
|
175
283
|
<tr>
|
|
176
|
-
<th>
|
|
177
|
-
<th
|
|
178
|
-
<th
|
|
284
|
+
<th>rv(color: :blue, texture: :rough)</th>
|
|
285
|
+
<th></th>
|
|
286
|
+
<th></th>
|
|
287
|
+
<th></th>
|
|
179
288
|
<th></th>
|
|
180
289
|
<th></th>
|
|
181
290
|
<th></th>
|
|
182
|
-
<th>✓</th>
|
|
183
291
|
</tr>
|
|
184
292
|
<tr>
|
|
185
|
-
<th>
|
|
186
|
-
<th
|
|
293
|
+
<th>rv(:color)</th>
|
|
294
|
+
<th>✓</th>
|
|
295
|
+
<th>✓</th>
|
|
296
|
+
<th>✓</th>
|
|
187
297
|
<th>✓</th>
|
|
188
298
|
<th></th>
|
|
189
299
|
<th></th>
|
|
300
|
+
</tr>
|
|
301
|
+
<tr>
|
|
302
|
+
<th>rv(:color, :texture)</th>
|
|
303
|
+
<th>Δ!</th>
|
|
304
|
+
<th>Δ!</th>
|
|
305
|
+
<th>Δ!</th>
|
|
306
|
+
<th>Δ!</th>
|
|
307
|
+
<th>Δ!</th>
|
|
190
308
|
<th></th>
|
|
191
|
-
<th>✓</th>
|
|
192
309
|
</tr>
|
|
310
|
+
</table>
|
|
311
|
+
|
|
312
|
+
### The Entropy Operator: .entropy()
|
|
313
|
+
|
|
314
|
+
<table>
|
|
193
315
|
<tr>
|
|
194
|
-
<th
|
|
316
|
+
<th></th>
|
|
317
|
+
<th>n/a</th>
|
|
318
|
+
<th>.given(:size)</th>
|
|
319
|
+
<th>.given(size: :small)</th>
|
|
195
320
|
<th>.given(size: :small, weight: :fat)</th>
|
|
196
|
-
<th
|
|
321
|
+
<th>.given(:size, weight: :fat)</th>
|
|
322
|
+
<th>.given(:size, :weight)</th>
|
|
323
|
+
</tr>
|
|
324
|
+
<tr>
|
|
325
|
+
<th>rv(color: :blue)</th>
|
|
326
|
+
<th></th>
|
|
327
|
+
<th></th>
|
|
328
|
+
<th></th>
|
|
197
329
|
<th></th>
|
|
198
330
|
<th></th>
|
|
199
331
|
<th></th>
|
|
200
|
-
<th>✓</th>
|
|
201
332
|
</tr>
|
|
202
333
|
<tr>
|
|
203
|
-
<th>
|
|
334
|
+
<th>rv(color: [:blue, :green])</th>
|
|
335
|
+
<th></th>
|
|
336
|
+
<th></th>
|
|
337
|
+
<th></th>
|
|
204
338
|
<th></th>
|
|
205
339
|
<th></th>
|
|
206
|
-
<th>✓</th>
|
|
207
|
-
<th>✓</th>
|
|
208
340
|
<th></th>
|
|
209
|
-
<th>✓</th>
|
|
210
341
|
</tr>
|
|
211
342
|
<tr>
|
|
212
|
-
<th>
|
|
213
|
-
<th
|
|
343
|
+
<th>rv(color: :blue, texture: :rough)</th>
|
|
344
|
+
<th></th>
|
|
345
|
+
<th></th>
|
|
346
|
+
<th></th>
|
|
347
|
+
<th></th>
|
|
214
348
|
<th></th>
|
|
349
|
+
<th></th>
|
|
350
|
+
</tr>
|
|
351
|
+
<tr>
|
|
352
|
+
<th>rv(:color)</th>
|
|
215
353
|
<th>✓</th>
|
|
216
354
|
<th>✓</th>
|
|
217
355
|
<th>✓</th>
|
|
218
356
|
<th>✓</th>
|
|
357
|
+
<th>✓</th>
|
|
358
|
+
<th></th>
|
|
219
359
|
</tr>
|
|
220
360
|
<tr>
|
|
221
|
-
<th>
|
|
222
|
-
<th
|
|
223
|
-
<th
|
|
361
|
+
<th>rv(:color, :texture)</th>
|
|
362
|
+
<th>✓</th>
|
|
363
|
+
<th>Δ!</th>
|
|
224
364
|
<th>✓</th>
|
|
365
|
+
<th>Δ!</th>
|
|
225
366
|
<th>✓</th>
|
|
226
367
|
<th></th>
|
|
227
|
-
<th>✓</th>
|
|
228
368
|
</tr>
|
|
369
|
+
</table>
|
|
370
|
+
|
|
371
|
+
### The Information Gain Operator: .infogain()
|
|
372
|
+
|
|
373
|
+
<table>
|
|
229
374
|
<tr>
|
|
230
|
-
<th
|
|
375
|
+
<th></th>
|
|
376
|
+
<th>n/a</th>
|
|
377
|
+
<th>.given(:size)</th>
|
|
378
|
+
<th>.given(size: :small)</th>
|
|
231
379
|
<th>.given(size: :small, weight: :fat)</th>
|
|
380
|
+
<th>.given(:size, weight: :fat)</th>
|
|
381
|
+
<th>.given(:size, :weight)</th>
|
|
382
|
+
</tr>
|
|
383
|
+
<tr>
|
|
384
|
+
<th>rv(color: :blue)</th>
|
|
385
|
+
<th></th>
|
|
386
|
+
<th></th>
|
|
387
|
+
<th></th>
|
|
388
|
+
<th></th>
|
|
232
389
|
<th></th>
|
|
233
|
-
<th>✓</th>
|
|
234
|
-
<th>✓</th>
|
|
235
390
|
<th></th>
|
|
236
|
-
<th>✓</th>
|
|
237
391
|
</tr>
|
|
238
392
|
<tr>
|
|
239
|
-
<th>
|
|
240
|
-
<th
|
|
393
|
+
<th>rv(color: [:blue, :green])</th>
|
|
394
|
+
<th></th>
|
|
395
|
+
<th></th>
|
|
396
|
+
<th></th>
|
|
397
|
+
<th></th>
|
|
241
398
|
<th></th>
|
|
242
399
|
<th></th>
|
|
243
|
-
<th>✓</th>
|
|
244
|
-
<th>✓</th>
|
|
245
|
-
<th>✓</th>
|
|
246
400
|
</tr>
|
|
247
401
|
<tr>
|
|
248
|
-
<th>
|
|
402
|
+
<th>rv(color: :blue, texture: :rough)</th>
|
|
249
403
|
<th></th>
|
|
250
404
|
<th></th>
|
|
251
|
-
<th
|
|
405
|
+
<th></th>
|
|
406
|
+
<th></th>
|
|
407
|
+
<th></th>
|
|
408
|
+
<th></th>
|
|
409
|
+
</tr>
|
|
410
|
+
<tr>
|
|
411
|
+
<th>rv(:color)</th>
|
|
412
|
+
<th></th>
|
|
252
413
|
<th>✓</th>
|
|
253
414
|
<th></th>
|
|
415
|
+
<th></th>
|
|
254
416
|
<th>✓</th>
|
|
417
|
+
<th></th>
|
|
255
418
|
</tr>
|
|
256
419
|
<tr>
|
|
257
|
-
<th>
|
|
258
|
-
<th
|
|
420
|
+
<th>rv(:color, :texture)</th>
|
|
421
|
+
<th></th>
|
|
422
|
+
<th></th>
|
|
423
|
+
<th></th>
|
|
424
|
+
<th></th>
|
|
259
425
|
<th></th>
|
|
260
|
-
<th>Δ!</th>
|
|
261
|
-
<th>Δ!</th>
|
|
262
426
|
<th></th>
|
|
263
|
-
<th>✓</th>
|
|
264
427
|
</tr>
|
|
428
|
+
</table>
|
|
429
|
+
|
|
430
|
+
### The Count Operator: .count()
|
|
431
|
+
|
|
432
|
+
<table>
|
|
265
433
|
<tr>
|
|
266
|
-
<th>Ps.rv(:color, :texture)</th>
|
|
267
|
-
<th>.given(size: :small)</th>
|
|
268
434
|
<th></th>
|
|
269
|
-
<th
|
|
435
|
+
<th>n/a</th>
|
|
436
|
+
<th>.given(:size)</th>
|
|
437
|
+
<th>.given(size: :small)</th>
|
|
438
|
+
<th>.given(size: :small, weight: :fat)</th>
|
|
439
|
+
<th>.given(:size, weight: :fat)</th>
|
|
440
|
+
<th>.given(:size, :weight)</th>
|
|
441
|
+
</tr>
|
|
442
|
+
<tr>
|
|
443
|
+
<th>rv(color: :blue)</th>
|
|
444
|
+
<th>✓</th>
|
|
445
|
+
<th>✓</th>
|
|
446
|
+
<th>✓</th>
|
|
447
|
+
<th>✓</th>
|
|
270
448
|
<th>✓</th>
|
|
271
|
-
<th></th>
|
|
272
449
|
<th>✓</th>
|
|
273
450
|
</tr>
|
|
274
451
|
<tr>
|
|
275
|
-
<th>
|
|
276
|
-
<th
|
|
277
|
-
<th
|
|
278
|
-
<th
|
|
279
|
-
<th
|
|
280
|
-
<th
|
|
452
|
+
<th>rv(color: [:blue, :green])</th>
|
|
453
|
+
<th>✓</th>
|
|
454
|
+
<th>✓</th>
|
|
455
|
+
<th>✓</th>
|
|
456
|
+
<th>✓</th>
|
|
457
|
+
<th>✓</th>
|
|
281
458
|
<th>✓</th>
|
|
282
459
|
</tr>
|
|
283
460
|
<tr>
|
|
284
|
-
<th>
|
|
285
|
-
<th
|
|
286
|
-
<th
|
|
287
|
-
<th
|
|
461
|
+
<th>rv(color: :blue, texture: :rough)</th>
|
|
462
|
+
<th>✓</th>
|
|
463
|
+
<th>✓</th>
|
|
464
|
+
<th>✓</th>
|
|
465
|
+
<th>✓</th>
|
|
466
|
+
<th>✓</th>
|
|
467
|
+
<th>✓</th>
|
|
468
|
+
</tr>
|
|
469
|
+
<tr>
|
|
470
|
+
<th>rv(:color)</th>
|
|
471
|
+
<th>✓</th>
|
|
472
|
+
<th>✓</th>
|
|
473
|
+
<th>✓</th>
|
|
474
|
+
<th>✓</th>
|
|
475
|
+
<th>✓</th>
|
|
476
|
+
<th>✓</th>
|
|
477
|
+
</tr>
|
|
478
|
+
<tr>
|
|
479
|
+
<th>rv(:color, :texture)</th>
|
|
480
|
+
<th>✓</th>
|
|
481
|
+
<th>✓</th>
|
|
482
|
+
<th>✓</th>
|
|
483
|
+
<th>✓</th>
|
|
288
484
|
<th>✓</th>
|
|
289
|
-
<th></th>
|
|
290
485
|
<th>✓</th>
|
|
291
486
|
</tr>
|
|
292
487
|
</table>
|
|
293
488
|
|
|
294
|
-
## Examples
|
|
295
|
-
|
|
296
|
-
There are examples of using Prolly to write learning algorithms.
|
|
297
|
-
|
|
298
|
-
- [Decision Tree](https://github.com/iamwilhelm/prolly/tree/master/examples/decision_tree)
|
|
299
|
-
|
|
300
|
-
### Probabilities
|
|
301
|
-
|
|
302
|
-
What is the probability there is a blue marble?
|
|
303
|
-
```ruby
|
|
304
|
-
# P(C = blue)
|
|
305
|
-
Ps.rv(color: :blue).prob
|
|
306
|
-
```
|
|
307
|
-
|
|
308
|
-
What is the joint probability there is a blue marble that also has a rough texture?
|
|
309
|
-
```ruby
|
|
310
|
-
# P(C = blue, T = rough)
|
|
311
|
-
Ps.rv(color: :blue, texture: :rough).prob
|
|
312
|
-
```
|
|
313
|
-
|
|
314
|
-
What is the probability a marble is small or med sized?
|
|
315
|
-
```ruby
|
|
316
|
-
# P(S = small, med)
|
|
317
|
-
Ps.rv(size: [:small, :med]).prob
|
|
318
|
-
```
|
|
319
|
-
|
|
320
|
-
What is the probability of a blue marble given that the marble is small?
|
|
321
|
-
```ruby
|
|
322
|
-
# P(C = blue | S = small)
|
|
323
|
-
Ps.rv(color: :blue).given(size: :small).prob
|
|
324
|
-
```
|
|
325
|
-
|
|
326
|
-
What is the probability of a blue marble and rough texture given that the marble is small?
|
|
327
|
-
```ruby
|
|
328
|
-
# P(C = blue, T = rough | S = small)
|
|
329
|
-
Ps.rv(color: :blue, texture: :rough).given(size: :small).prob
|
|
330
|
-
```
|
|
331
|
-
|
|
332
|
-
### Probability density functions
|
|
333
|
-
|
|
334
|
-
Probability density for a random variable.
|
|
335
|
-
```ruby
|
|
336
|
-
Ps.rv(:color).pdf
|
|
337
|
-
```
|
|
338
|
-
|
|
339
|
-
Probability density for a conditional random variable.
|
|
340
|
-
```ruby
|
|
341
|
-
Ps.rv(:color).given(size: :small).pdf
|
|
342
|
-
```
|
|
343
|
-
|
|
344
|
-
### Entropy
|
|
345
|
-
|
|
346
|
-
Entropy of the RV color.
|
|
347
|
-
```ruby
|
|
348
|
-
# H(C)
|
|
349
|
-
Ps.rv(:color).entropy
|
|
350
|
-
```
|
|
351
|
-
|
|
352
|
-
Entropy of color given the marble is small
|
|
353
|
-
```ruby
|
|
354
|
-
# H(C | S = small)
|
|
355
|
-
Ps.rv(:color).given(size: :small).entropy
|
|
356
|
-
```
|
|
357
|
-
|
|
358
|
-
### Information Gain
|
|
359
|
-
|
|
360
|
-
Information gain of color and size.
|
|
361
|
-
```ruby
|
|
362
|
-
# IG(C | S)
|
|
363
|
-
Ps.rv(:color).given(:size).infogain
|
|
364
|
-
```
|
|
365
|
-
### Counts
|
|
366
|
-
|
|
367
|
-
At the base of all the probabilities are counts of stuff.
|
|
368
|
-
```ruby
|
|
369
|
-
Ps.rv(color: :blue).count
|
|
370
|
-
```
|
|
371
|
-
|
|
372
|
-
```ruby
|
|
373
|
-
Ps.rv(:color).given(:size).count
|
|
374
|
-
```
|
|
375
|
-
|
|
376
489
|
## Stores
|
|
377
490
|
|
|
378
491
|
Prolly can use different stores to remember the prior event data from which it
|
data/lib/prolly/ps.rb
CHANGED
|
@@ -60,9 +60,9 @@ module Prolly
|
|
|
60
60
|
|
|
61
61
|
def_delegators :@storage, :reset, :add, :count, :rand_vars, :uniq_vals, :import
|
|
62
62
|
|
|
63
|
-
def initialize
|
|
64
|
-
#@storage = Storage::
|
|
65
|
-
@storage = Storage::
|
|
63
|
+
def initialize(storage = nil)
|
|
64
|
+
#@storage = Storage::Mongodb.new()
|
|
65
|
+
@storage = Storage::Rubylist.new()
|
|
66
66
|
#@storage = Storage::Redis.new()
|
|
67
67
|
end
|
|
68
68
|
|
|
@@ -22,18 +22,16 @@ module Prolly
|
|
|
22
22
|
|
|
23
23
|
def count(rvs, options = {})
|
|
24
24
|
reload = options[:reload] || false
|
|
25
|
-
start_time = Time.now
|
|
26
25
|
if rvs.kind_of?(Array)
|
|
27
26
|
value = @data.count { |e| rvs.all? { |rv| e.has_key?(rv) } }
|
|
28
27
|
elsif rvs.kind_of?(Hash)
|
|
29
28
|
value = @data.count { |e|
|
|
30
29
|
rvs.map { |rkey, rval|
|
|
31
30
|
vals = rval.kind_of?(Array) ? rval : [rval]
|
|
32
|
-
vals.include?(e[rkey])
|
|
31
|
+
vals.include?(e[rkey])
|
|
33
32
|
}.all?
|
|
34
33
|
}
|
|
35
34
|
end
|
|
36
|
-
elapsed = Time.now - start_time
|
|
37
35
|
return value
|
|
38
36
|
end
|
|
39
37
|
|