SHOW:
|
|
- or go back to the newest paste.
1 | - | # Calcualte stats for Wheel of Time series |
1 | + | #!/usr/bin/env python |
2 | - | # Each row is the number of 5,4,3,2,1 rations for a book in the series |
2 | + | # -*- coding: utf-8 -*- |
3 | - | # Collected from goodreads.com, amazon.com and amazon.co.uk |
3 | + | """ |
4 | Calcualte stats for Wheel of Time series | |
5 | Each row is the number of 5,4,3,2,1 rations for a book in the series | |
6 | Collected from goodreads.com, amazon.com and amazon.co.uk | |
7 | """ | |
8 | ||
9 | goodreads = [ | |
10 | [45132, 32768, 15394, 4462, 2287], # 1 | |
11 | [-1, -1, -1, -1, -1], # 2 | |
12 | [31992, 26753, 11780, 2265, 795], # 3 | |
13 | [24523, 21922, 10282, 1747, 430], # 4 | |
14 | [16119, 16903, 9131, 1749, 371], # 5 | |
15 | [12290, 12332, 7866, 1750, 399], # 6 | |
16 | [9925, 11312, 8258, 2068, 451], # 7 | |
17 | [8841, 9805, 8043, 2511, 542], # 8 | |
18 | [8891, 9205, 7464, 2438, 622], # 9 | |
19 | [7411, 7549, 6040, 2272, 864], # 10 | |
20 | [8959, 8456, 4699, 1202, 321], # 11 | |
21 | [14569, 9046, 3169, 676, 363], # 12 | |
22 | [14147, 7671, 2337, 451, 292]] # 13 | |
23 | ||
24 | amazoncom = [ | |
25 | [1279, 404, 188, 138, 157], # 1 | |
26 | [393, 139, 37, 13, 15], # 2 | |
27 | [292, 122, 35, 23, 17], # 3 | |
28 | [275, 93, 39, 17, 19], # 4 | |
29 | [204, 97, 54, 27, 24], # 5 | |
30 | [227, 83, 55, 35, 35], # 6 | |
31 | [335, 187, 116, 73, 54], # 7 | |
32 | [361, 330, 379, 352, 466], # 8 | |
33 | [331, 302, 191, 184, 185], # 9 | |
34 | [199, 146, 253, 396, 1581], # 10 | |
35 | [184, 155, 108, 81, 111], # 11 | |
36 | [540, 149, 31, 22, 14], # 12 | |
37 | [1628, 346, 150, 62, 358]] # 13 | |
38 | ||
39 | amazoncouk = [ | |
40 | [144, 47, 20, 19, 20], # 1 | |
41 | [33, 25, 4, 0, 1], # 2 | |
42 | [46, 25, 8, 0, 5], # 3 | |
43 | [32, 12, 7, 0, 1], # 4 | |
44 | [21, 17, 5, 5, 4], # 5 | |
45 | [25, 10, 10, 1, 2], # 6 | |
46 | [19, 12, 2, 0, 0], # 7 | |
47 | [20, 21, 22, 16, 19], # 8 | |
48 | [44, 33, 26, 9, 10], # 9 | |
49 | [20, 18, 32, 60, 76], # 10 | |
50 | [38, 42, 17, 6, 5], # 11 | |
51 | [125, 22, 6, 3, 1], # 12 | |
52 | [96, 31, 8, 8, 5]] # 13 | |
53 | ||
54 | import math | |
55 | import numpy as np | |
56 | from matplotlib import pyplot as plot | |
57 | ||
58 | ||
59 | def histo_stat(haxis, histo): | |
60 | """ | |
61 | Parameters: haxis - x axis, histo - weight for each entry in axis | |
62 | - | return 0, 0, [] |
62 | + | |
63 | """ | |
64 | sumw = sum(h for h in histo) | |
65 | if sumw < 0: | |
66 | return 0, 0, [0] * len(haxis) | |
67 | ihisto = zip(histo, haxis) | |
68 | mean = 1. / sumw * sum(w * x for w, x in ihisto) | |
69 | std = math.sqrt(1. / sumw * sum(i[0] * (i[1] - mean) ** 2 for i in ihisto)) | |
70 | - | def run(name, data, color): |
70 | + | |
71 | return mean, std, frac | |
72 | ||
73 | ||
74 | def run(name, data): | |
75 | """ | |
76 | Paramters: name - datasetname, data - table with histgrams | |
77 | """ | |
78 | fig = plot.figure(1) | |
79 | - | # Special case with no data |
79 | + | |
80 | - | if name == "Goodreads.com" and i == 1: |
80 | + | |
81 | stats = [] | |
82 | percs = [] | |
83 | for i, histo in enumerate(data): | |
84 | mean, std, frac = histo_stat(haxis, histo) | |
85 | - | stats = np.array(stats) |
85 | + | if mean == 0: # Special case with no data |
86 | mean = 4.15 | |
87 | - | xaxis = range(1, len(data)+1) |
87 | + | |
88 | - | plot.errorbar(xaxis, stats[:, 0], xerr=0.5, yerr=stats[:, 1], color=color, fmt='-o') |
88 | + | |
89 | - | plot.fill_between(xaxis, stats[:, 0] - stats[:, 1], stats[:, 0] + stats[:, 1], facecolor=color, alpha=0.3) |
89 | + | percs += [[100 * fi for fi in frac]] |
90 | print "{:>2}: {:.2f} +- {:.2f} [{}]".format(i + 1, mean, std, perc) | |
91 | - | run("Amazon.com", amazoncom, "red") |
91 | + | return np.array(stats), np.array(percs) |
92 | - | run("Amazon.co.uk", amazoncouk, "blue") |
92 | + | |
93 | - | run("Goodreads.com", goodreads, "green") |
93 | + | |
94 | def plot_stats(stats, color): | |
95 | """ | |
96 | Plot statistics, can superimpose | |
97 | """ | |
98 | nx = stats.shape[0] | |
99 | ax = plot.subplot(1, 1, 1, xlim=(1, 13)) | |
100 | xaxis = range(1, nx + 1) | |
101 | plot.errorbar( | |
102 | xaxis, stats[:, 0], xerr=0.5, yerr=stats[:, 1], color=color, fmt='-o') | |
103 | plot.fill_between(xaxis, stats[:, 0] - stats[:, 1], stats[ | |
104 | :, 0] + stats[:, 1], facecolor=color, alpha=0.3) | |
105 | plot.ylabel("score") | |
106 | plot.xlabel("book") | |
107 | ||
108 | ||
109 | def plot_percs(isub, percs, name): | |
110 | """ | |
111 | Plot percentages as image to a give subfig | |
112 | """ | |
113 | plot.subplot(3, 1, isub) | |
114 | im = plot.imshow( | |
115 | percs.T, extent=(1, percs.T.shape[1], 1, percs.T.shape[0]), | |
116 | interpolation="nearest", clim=(0, 100)) | |
117 | plot.title(name) | |
118 | plot.ylabel("score") | |
119 | plot.xlabel("book") | |
120 | return im | |
121 | ||
122 | stats1, percs1 = run("Amazon.com", amazoncom) | |
123 | stats2, percs2 = run("Amazon.co.uk", amazoncouk) | |
124 | stats3, percs3 = run("Goodreads.com", goodreads) | |
125 | ||
126 | fig1 = plot.figure(1) | |
127 | plot_stats(stats1, "red") | |
128 | plot_stats(stats2, "blue") | |
129 | plot_stats(stats3, "green") | |
130 | ||
131 | fig2 = plot.figure(2) | |
132 | plot_percs(1, percs1, "Amazon.com") | |
133 | plot_percs(2, percs2, "Amazon.co.uk") | |
134 | im = plot_percs(3, percs3, "Goodreads.com") | |
135 | cax = fig2.add_axes([0.8, 0.1, 0.03, 0.85]) | |
136 | fig2.colorbar(im, cax=cax, label="Percentage") | |
137 | fig2.tight_layout() | |
138 | plot.show() |