| 查看: 562 | 回复: 4 | ||
[求助]
帮忙解释加下划线的四句code，谢谢
|
|
Sample Text %Procedure for the likelihood% function f=lpr(x) global de dr pres uptop nd n prp prp2 ndr qp; % Initial Quality Perception; % True Quality Perception; % Detailing Signal Variance; % Initial Quality Variance iqp=x(1); tqp=x(2); dsv=exp(x(3)); iqv=1; ndr=zeros(n*6,nd); qp=ndr; for i=1:nd %Distribution of Signals for each time period% Sample Text %Mean quality perceptions for each time period% qp(:,i)=((iqp/iqv)+(ndr(:,i)/dsv))./((1/iqv)+(de/dsv)); end %Prescription probability% prp=exp(qp)./(1+exp(qp)); %No prescription probability% nprp=1./(1+exp(qp)); %Choose appropriate probability or 1 depending upon behavior and number of periods physician active in data% prp2=zeros(n*6,nd); for i=1:nd prp2(:,i)=((prp(:,i).*pres)+(nprp(:,i).*(1-pres))).^uptop; end %Reshaping to compute likelihood% Sample Text |
回复此楼» 猜你喜欢
今年的WR进展到哪一步了？
已经有8人回复
-
时间戳他又来了
已经有17人回复
-
系统今天又提示维护了，估计离放榜不远了
已经有9人回复
-
产物和副产物价值比较
已经有4人回复
-
这个自发加氧反应的机理是什么？
已经有7人回复
-
J-J-W不为人知的一面
已经有19人回复
-
没消息就是被刷了呗
已经有14人回复
-
微信指数没变化，科研之友没阅读
已经有8人回复
-
这种情况还有戏吗
已经有11人回复
-
同事接到电话了，我却没有
已经有4人回复
|
ndr(:,i)=tqp*de+(sqrt(dsv)*sqrt(de)).*dr(:,i); prp2=reshape(prp2',n*nd,6); prp2=mean(reshape((prod(prp2',1))',nd,n),1); f=-sum(sum(log(prp2))); %Procedure for the likelihood% function f=lpr(x) global de dr pres uptop nd n prp prp2 ndr qp; % Initial Quality Perception; % True Quality Perception; % Detailing Signal Variance; % Initial Quality Variance iqp=x(1); tqp=x(2); dsv=exp(x(3)); iqv=1; ndr=zeros(n*6,nd); qp=ndr; for i=1:nd %Distribution of Signals for each time period% ndr(:,i)=tqp*de+(sqrt(dsv)*sqrt(de)).*dr(:,i); %Mean quality perceptions for each time period% qp(:,i)=((iqp/iqv)+(ndr(:,i)/dsv))./((1/iqv)+(de/dsv)); end %Prescription probability% prp=exp(qp)./(1+exp(qp)); %No prescription probability% nprp=1./(1+exp(qp)); %Choose appropriate probability or 1 depending upon behavior and number of periods physician active in data% prp2=zeros(n*6,nd); for i=1:nd prp2(:,i)=((prp(:,i).*pres)+(nprp(:,i).*(1-pres))).^uptop; end %Reshaping to compute likelihood% prp2=reshape(prp2',n*nd,6); prp2=mean(reshape((prod(prp2',1))',nd,n),1); f=-sum(sum(log(prp2))); |
2楼2014-02-09 03:55:51
3楼2014-02-09 03:57:07
4楼2014-02-09 12:31:17
5楼2014-02-09 13:06:26