Mercurial > hg > cc > cirrus_work

comparison bin/spearman.py @ 31:e7c8e64c2fdd

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
get multi-ranking done right
author Henry S. Thompson <ht@inf.ed.ac.uk>
date Thu, 17 Nov 2022 13:51:19 +0000
parents c73ec9deabbe
children 91741bf3ab51
comparison
equal deleted inserted replaced
30:c73ec9deabbe 31:e7c8e64c2fdd
43 plt.show() 43 plt.show()
44 sm.qqplot(x[xworst], line='s') 44 sm.qqplot(x[xworst], line='s')
45 plt.gca().set_title('Worst segment (most variance): %s'%xworst) 45 plt.gca().set_title('Worst segment (most variance): %s'%xworst)
46 plt.show() 46 plt.show()
47 47
48 def plot_x(): 48 def plot_x(block=True):
49 plt.plot([xd[i].mean for i in range(100)],'bx',label='Mean of rank correlation of each segment x all other segments') 49 plt.plot([xd[i].mean for i in range(100)],'bx',label='Mean of rank correlation of each segment x all other segments')
50 plt.plot([0,99],[xm,xm],'b',label='Mean of segment x segment means') 50 plt.plot([0,99],[xm,xm],'b',label='Mean of segment x segment means')
51 plt.plot(all,'rx',label='Rank correlation of segment x whole crawl') 51 plt.plot(all,'rx',label='Rank correlation of segment x whole crawl')
52 plt.plot([0,99],[all_m,all_m],'r',label='Mean of segment x whole crawl') 52 plt.plot([0,99],[all_m,all_m],'r',label='Mean of segment x whole crawl')
53 plt.axis([0,99,0.8,1.0]) 53 plt.axis([0,99,0.8,1.0])
54 plt.legend(loc='best') 54 plt.legend(loc='best')
55 plt.grid(True) 55 plt.grid(True)
56 plt.show() 56 plt.show(block=block)
57 57
58 def hist(): 58 def hist():
59 sdd=[(i,xm-(i*xsd)) for i in range(-2,3)] 59 sdd=[(i,xm-(i*xsd)) for i in range(-2,3)]
60 fig,hax=plt.subplots() # Thanks to https://stackoverflow.com/a/7769497 60 fig,hax=plt.subplots() # Thanks to https://stackoverflow.com/a/7769497
61 sdax=hax.twiny() 61 sdax=hax.twiny()
77 return i+1 77 return i+1
78 78
79 def ranks(): 79 def ranks():
80 # Combine segment measures: 80 # Combine segment measures:
81 # segID,rank corr. wrt all,inverse variance, mean cross rank corr.,first disagreement 81 # segID,rank corr. wrt all,inverse variance, mean cross rank corr.,first disagreement
82 return np.array([i,all[i],1.0/xd[i].variance,xd[i].mean,first_diff(ranks[i])]) 82 # convert to ranks, smallest value == highest rank
83 all_ranked=stats.rankdata(-all,method='average') # invert since
84 # large corr is good
85 x_variance_ranked=stats.rankdata([xd[i].variance for i in range(100)])
86 # small corr variance is good
87 x_mean_ranked=stats.rankdata([-(xd[i].mean) for i in range(100)])
88 # invert since
89 # large mean corr is good
90 fd_ranked=stats.rankdata([-first_diff(x_ranks[i]) for i in range(100)])
91 # invert since
92 # large first diff is good
93 return np.array([[i,
94 all_ranked[i],
95 x_variance_ranked[i],
96 x_mean_ranked[i],
97 fd_ranked[i]] for i in range(100)])
83 98
84 counts=loadtxt(sys.argv[1]+".csv",delimiter=',') 99 counts=loadtxt(sys.argv[1]+".csv",delimiter=',')
85 # "If axis=0 (default), then each column represents a variable, with 100 # "If axis=0 (default), then each column represents a variable, with
86 # observations in the rows" 101 # observations in the rows"
87 # So each column is a sequence of counts, for whole crawl in column 0 102 # So each column is a sequence of counts, for whole crawl in column 0
99 xd=[stats.describe(x[i]) for i in range(100)] 114 xd=[stats.describe(x[i]) for i in range(100)]
100 xs=stats.describe(np.array([xd[i].mean for i in range(100)])) 115 xs=stats.describe(np.array([xd[i].mean for i in range(100)]))
101 xm=xs.mean 116 xm=xs.mean
102 xsd=np.sqrt(xs.variance) 117 xsd=np.sqrt(xs.variance)
103 118
104 ranks=[stats.rankdata(-counts[:,i],method='average') for for i in range(1,100)] 119 x_ranks=[stats.rankdata(-counts[:,i],method='average') for i in range(1,101)]
105 120
106 ### I need to review rows, e.g. counts[0] is an array of 101 counts 121 ### I need to review rows, e.g. counts[0] is an array of 101 counts
107 ### for the most common label in the complete crawl, 122 ### for the most common label in the complete crawl,
108 ### from the complete crawl and all the segments 123 ### from the complete crawl and all the segments
109 ### versus columns, e.g. counts[:,0] is an array of 100 decreasing counts 124 ### versus columns, e.g. counts[:,0] is an array of 100 decreasing counts