From 484adfcda822f25153cce8a1e027f266382a3f59 Mon Sep 17 00:00:00 2001 From: art Date: Wed, 2 Oct 2019 20:56:06 -0500 Subject: [PATCH 1/4] stuff --- data analysis/analysis/analysis.py | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/data analysis/analysis/analysis.py b/data analysis/analysis/analysis.py index d018c524..eea0425e 100644 --- a/data analysis/analysis/analysis.py +++ b/data analysis/analysis/analysis.py @@ -295,6 +295,11 @@ def regression_engine(device, inputs, outputs, args, loss = torch.nn.MSELoss(), return regressions +#@jit TODO: determine jit type +def elo(starting_score): + + + @jit(forceobj=True) def r_squared(predictions, targets): # assumes equal size inputs From acdcb42e6d427d3ffeb9f71f6dc938b8299d9e72 Mon Sep 17 00:00:00 2001 From: art Date: Wed, 2 Oct 2019 20:57:09 -0500 Subject: [PATCH 2/4] quick tests --- data analysis/analysis/analysis.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/data analysis/analysis/analysis.py b/data analysis/analysis/analysis.py index eea0425e..cc8feedd 100644 --- a/data analysis/analysis/analysis.py +++ b/data analysis/analysis/analysis.py @@ -296,9 +296,9 @@ def regression_engine(device, inputs, outputs, args, loss = torch.nn.MSELoss(), return regressions #@jit TODO: determine jit type -def elo(starting_score): +def elo(starting_score, observed, N, K): - + pass @jit(forceobj=True) def r_squared(predictions, targets): # assumes equal size inputs From 8801a300c4253f7fa26d7abee4bbe7b476d99841 Mon Sep 17 00:00:00 2001 From: art Date: Thu, 3 Oct 2019 10:42:05 -0500 Subject: [PATCH 3/4] analysis.py v 1.1.2.002 --- .../__pycache__/analysis.cpython-36.pyc | Bin 16024 -> 16290 bytes data analysis/analysis/analysis.py | 13 +++++++++---- 2 files changed, 9 insertions(+), 4 deletions(-) diff --git a/data analysis/analysis/__pycache__/analysis.cpython-36.pyc b/data analysis/analysis/__pycache__/analysis.cpython-36.pyc index 16f3a706285cc0b1501163b5e2e09c019f9ddb7c..c61d176c9f268d98f91832a330b5a468b6b82edd 100644 GIT binary patch delta 2170 zcmb_dTTGlq6rR~-m*rY67PbP*(sKI?EDNPjZe=S%%Uv!t@TYWLcK&X+ghuW7um_BLECdMWvM(vBnOZ1`9IRoylP=gQdX1_V} z&7AXpXXc!L-(0w|=-iiiJyLr62gc|7*#2&`PgZti)=UB zvqIE7&CIN9*D7VjEMDOVKb<{*3z)V5=qG{v&%woh}JDy*;07k{v5F)4OJYA+Zh78Np zO6HAuB%9=tJhEV#Hwpd5eqJs1f1fNsF><>Yc#!~ghaU3qkU!va>)xOctC-dFe^Tm> zf#CFXP>=0VEx`$0<7YJ%q5Tosqgu4tY0Vqb7|+4L5^{wVF^Uta5cjMnh#4^y8~711 zg3HzdzZSS?4b@Gbj2qwOiNMWGQI*rbT#!b`AL{JI$}J_JK0rKTZauEIjecXSX=lsZj= zV*VPuRXRs6g>%a6)HaHHH|1t=yZr%w$nCxYBjxAICd8JAnQPOUC&Z88{#}xp@`R?& zOmdHyh=cdS!}1FHE-1?345T=^afaZBB_*G}4VRUB4Q?dz5%AU&vy8=`iLx=*He5Gw zFD{MBGsIw5;nTe@*U8u#uArVsqVD%i1^sM2*a_tSR)w|?;iL@V7iJhPRdkod9wf6d z1Ok!7;f*k-&>TM=Gx*tKU|Evd_o?{@@Bwm2t)9Y@rN{l6A-l#gijsGi2-fK zJr$abYog!aLPOuqSQAN|akn)h{2qjp4lGmZht0&h@t?&gD5Zq1A$bUC*(5xC;Rr6Tfy@Yo{C$e@wNmgg2;MEe$(x`kOwkku8nC7t9y6q)dS+&o z7l73;!-6so+Mb-Uj~S&>l~eXJlT`LfrB+1MF}qZ%b4opPNX3#<8kkcm)|}#EjZ(4o zS{3GII~EkSV_v5j*sfv4;h8CEb~n3Q+#Zi7(vCkud8S#n`XJoEpG#kdTR8jB27HKP zW#7Xp2Fg=#7jKs@!54V4;#b(n95nTTmTQ?Pu5bE z?xRsGoUG_2(Mn(w;L!1cT}Gj_V8F`2i_6v!bYh7u1cz}_U?B5}ZEYJ2;8Lww9S|GS z_)+a8Ju!Y1N9*d;qk=BrT-~=2#3{Q+4T|U2@UnddE@mq0tpFqFb5z0G_@v`i`_p1^ zUC*PZqmdLpK}jm55{sl_XQp{bJUF)T@x=xwETOBRcNZy~qIj4fiq=-M+9~KIywb4d z9V4-dph|cQVe#Qh+h+a+etj{NlLV zq2FTieZ1=a;J_3m$eHA)=<#_1nVwGNd8&Ydc5TUa?`yt;E6vMY|5$;v;H7MTuYlBPf>`dxKf~3Q zz{B^dB}^O1ca<2KL#=;g@ArRMUK5rr_+G0KSKH?9WBni<`=6|D2wD c=SVHV-~5knr`J(YS^*sB>e=bvtrq_M30>=u+yDRo diff --git a/data analysis/analysis/analysis.py b/data analysis/analysis/analysis.py index cc8feedd..d0ce438d 100644 --- a/data analysis/analysis/analysis.py +++ b/data analysis/analysis/analysis.py @@ -7,10 +7,13 @@ # current benchmark of optimization: 1.33 times faster # setup: -__version__ = "1.1.2.001" +__version__ = "1.1.2.002" # changelog should be viewed using print(analysis.__changelog__) __changelog__ = """changelog: +1.1.2.002L + - added elo() + - elo() has bugs to be fixed 1.1.2.001: - readded regrression import 1.1.2.000: @@ -295,10 +298,12 @@ def regression_engine(device, inputs, outputs, args, loss = torch.nn.MSELoss(), return regressions -#@jit TODO: determine jit type -def elo(starting_score, observed, N, K): +@jit(nopython=True) +def elo(starting_score, opposing_scores, observed, N, K): - pass + expected = 1/(1+10**((np.array(opposing_scores) - starting_score)/N)) + + return starting_score + K*(np.sum(expected) - np.sum(observed)) @jit(forceobj=True) def r_squared(predictions, targets): # assumes equal size inputs From b6299ce3972a464eb170e9f9e46165cca0880486 Mon Sep 17 00:00:00 2001 From: art Date: Thu, 3 Oct 2019 10:48:56 -0500 Subject: [PATCH 4/4] analysis.py v 1.1.2.003 --- data analysis/analysis/analysis.py | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/data analysis/analysis/analysis.py b/data analysis/analysis/analysis.py index d0ce438d..f9f77603 100644 --- a/data analysis/analysis/analysis.py +++ b/data analysis/analysis/analysis.py @@ -7,11 +7,13 @@ # current benchmark of optimization: 1.33 times faster # setup: -__version__ = "1.1.2.002" +__version__ = "1.1.2.003" # changelog should be viewed using print(analysis.__changelog__) __changelog__ = """changelog: -1.1.2.002L +1.1.2.003: + - fixed elo() +1.1.2.002: - added elo() - elo() has bugs to be fixed 1.1.2.001: @@ -303,7 +305,7 @@ def elo(starting_score, opposing_scores, observed, N, K): expected = 1/(1+10**((np.array(opposing_scores) - starting_score)/N)) - return starting_score + K*(np.sum(expected) - np.sum(observed)) + return starting_score + K*(np.sum(observed) - np.sum(expected)) @jit(forceobj=True) def r_squared(predictions, targets): # assumes equal size inputs