MATLAB无法识别函数变量（BP神经网络相关），基础拟合

Question

MATLAB课BP神经网络尝试拟合函数，但是一直报错
以下为代码
P=-1:.1:1;
T=[-.9602 -.5770 -.0729 .3771 .6405 .6600 .4609 ...
.1336 -.2013 -.4344 -.5000 -.3930 -.1647 .0988 ...
.3072 .3960 .3449 .1816 -.0312 -.2189 -.3201];
plot(P,T,'+');
title('Training Vectors');
xlabel('Input Vector P');
ylabel('Target Vector T');
net=newff(P,T,5,{'tansig' 'purelin'},'traingd','learngd','sse');%在这里报错，内容为“MATLAB函数或变量 'newff' 无法识别”
me=8000;
net.trainParam.show=10;
net.trainParam.goal=0.02;
net.trainParam.lr=0.01;
A=sim(net,P);
sse=sumsqr(T-A);
for i=1:me/100 if sse<net.trainparam.goal,i=i-1;break,end
net.trainParam.epochs=100;
[net,tr]=train(net,P,T);
trp((1+100*(i-1)):(max(tr.epoch)+100*(i-1)))=tr.perf(1:max(tr.epoch));
A=sim(net,P);
sse=sumsqr(T-A);
plot(P,T,'+');
hold on
plot(P,A)
hold off
pause(1)
end
message=sprintf('Traingd,Epoch%%g/%g,SSE%%g\n',me);
fprintf(message,(max(tr.epoch)+100*(i-1)),sse)
plot(trp)
[i,j]=size(trp);
hold on
plot(1:j,net.trainParam.goal,'r--')
hold off
title('Error Signal')
xlabel('epoch')
ylabel('Error')
p=0.5;
a=sim(net,p)
echo off