Re: [問題] for迴圈加速

看板MATLAB作者jack10313 (彩色筆)時間9年前 (2015/06/17 02:16)推噓4(4推 0噓 7→)

留言11則, 4人參與討論串2/2 (看更多)

MAC OS X更新，MALTAB突然不能用害我剛剛才更新好基本上減少for迴圈數量，用矩陣取代 meshgrid是個不錯的func 解析度開太高也會遇到記憶體不足問題電腦太爛我就不嘗試了所以我尋找的條件就放寬不知道有沒有符合需求這是算微帶線嗎，我看不懂看你也是交大便宜收500就好 XDDD source code ------------------------------- clear all close all clc % Coding by Sheng-Hsun %---2015/06/16 copyleft --- %parameters e1 = 1; e2 = -62 -100j; vp=3E8; %main function [Reff Ieff lambda] = meshgrid(-62:1:1, -100:1:0, 2:1:20); k0 = 2*pi./lambda; %u1 theta1 = atan( (-Ieff)./(1-Reff) ); alpha1 = k0.*((1-Reff).^2+(Ieff).^2).^0.25.*sin(theta1/2); beta1 = k0.*((1-Reff).^2+(Ieff).^2).^0.25.*cos(theta1/2); %u2 theta2 = atan((100+Ieff)./(62+Reff)); alpha2 = k0.*((62+Reff).^2+(100+Ieff).^2).^0.25.*sin(theta2/2); beta2 = k0.*((62+Reff).^2+(100+Ieff).^2).^0.25.*cos(theta2/2); Z1 = (2*pi*vp)./((alpha1+beta1).*lambda); Z2 = (2*pi*vp)./((alpha2+beta2).*lambda); Q = ((Z1./Z2)+(Z2./Z1)).*0.5; t1 = cos(alpha1).*cosh(beta1).*cos(alpha2).*cosh(beta2); t2 = sin(alpha1).*sinh(beta1).*sin(alpha2).*sinh(beta2); t3 = sin(alpha1).*cosh(beta1).*sin(alpha2).*cosh(beta2); t4 = cos(alpha1).*sinh(beta1).*cos(alpha2).*sinh(beta2); temp1 = t1 + t2 + Q.*(t3+t4); Qur = cos(alpha1).*cosh(beta1).*sin(alpha2).*sinh(beta2); Qui = sin(alpha1).*sinh(beta1).*cos(alpha2).*cosh(beta2); Qdr = sin(alpha1).*cosh(beta1).*cos(alpha2).*sinh(beta2); Qdi = cos(alpha1).*sinh(beta1).*sin(alpha2).*cosh(beta2); temp2 = (Qur-Qui) - Q.*(Qdr-Qdi); err = 2E1; idx1 = find (abs(temp1-1) <= err); idx2 = find (abs(temp2) <= err) ; idx = intersect(idx1, idx2) Re = Reff(idx); Im = Ieff(idx); L = lambda(idx); plot3(L, Re, Im); ※ 引述《leo10843 (這不是末日!!)》之銘言： : 大家好，由於程式底子太差了，所以只能用最笨的方法，直接寫三個for迴圈QQ : 就算讓電腦一直掛著跑，發現還是跑不出來結果。 : 上網爬了些文章，發現不是沒有幫助就是底子太差看不懂@@ : 在此附上程式碼，https://goo.gl/oc4b7R : 請高手們幫忙看程式碼可以改哪裡 : 如果真的讓我跑出結果，奉上1000P幣(稅前) : P.S 如果幫助太踴躍，我只能挑出一位給P幣，太窮了XDD -- ※ 發信站: 批踢踢實業坊(ptt.cc), 來自: 140.113.125.75 ※ 文章網址: https://www.ptt.cc/bbs/MATLAB/M.1434478586.A.80B.html

→

jatj

06/17 11:30, , 1^F

06/17 11:30, 1^F

→

lavifoxbat

06/17 12:58, , 2^F

06/17 12:58, 2^F

推

profyang

06/17 13:29, , 3^F

06/17 13:29, 3^F