Using map2mat for eof analysis!

2017 3 13
3 个回答
更新时间：2017 5 16
4 次查看（30 天）

0 个投票

I have a matrix mm_r of the size 361*361*36; where 36 represent the number of years
i need to make it of size 36*130321
% Get EOFs:
G = map2mat(ones(size(mm_r,3),size(mm_r,2)),mm_r);
  • *the error message is Attempted to access F(37,1); index out of bounds because size(F)=[36,361].
Error in map2mat (line 30) if F(ix,iy)>0**

2 个评论
显示 无 隐藏 无

Walter Roberson
Walter Roberson 2017-3-13
Is mat2map from the pcatool File Exchange contribution?
Sophia
Sophia 2017-3-15
Yes, it is from pcatool file excahnge.

请先登录，再进行评论。

请先登录，再回答此问题。

回答（3 个）

Chad Greene
Chad Greene 2017-3-13

0 个投票

Hi Sophia,
Be careful, because map2mat assumes the first dimension is time. For normal climate data (lon*lat*time or lat*lon*time) it's probably easier to use eof, which does all the reshaping for you.

6 个评论
显示 4更早的评论 隐藏 4更早的评论

Sophia
Sophia 2017-3-13
I did saw the eof before, but i also saw you mentioned there about pcatool so i moved back to pcatool
Chad Greene
Chad Greene 2017-3-13
My eof function is little more than a wrapper for PCATool's caleof function to make it more user friendly.
Sophia
Sophia 2017-3-15
So i used this instead
[eofmap,pc] = eof(mm_r,1);
I cannot understand the result, honestly i am not even sure what to expect as a result.
Chad Greene
Chad Greene 2017-3-15
Hi Sophia,
I cannot understand your result either given the information you have provided.
Sophia
Sophia 2017-5-16
编辑：Sophia 2017-5-16
Hi
I used your eof technique as well..Somehow, does not matter what i choose as N, I mean 2 or 3 the total of all the eof is 99.9%, is there a reason for that or i am missing something..take a look at the codes below
Sophia
Sophia 2017-5-16
I changed it to 36, it seems somehow better now. is it like that ??

请先登录，再进行评论。

Sophia
Sophia 2017-5-16

0 个投票

** the total is 99.9% **
clear all; clc;
data= load('north_x_y_lat_lon');
datacoord = reshape(data, 361,361,4);
lat = squeeze(datacoord(:,:,3));
long = squeeze(datacoord(:,:,4));
rlong = long*pi/180.;
rlat = lat*pi/180.;
load nsidc_sid_1979_2014.mat
N1=2;
[eof_maps,pc,expv] = eof(t_u,2);
[eof_maps1,pc1,expv1] = eof(t_v,2);
figure(1);clf;iw=1;jw=N1+2;
set(gcf,'MenuBar','none');
for i=1:iw*jw
if i<= iw*jw-2
C1 = squeeze(eof_maps(:,:,i));
C2 = squeeze(eof_maps1(:,:,i));
pc_u = pc';
pc_v = pc1';
subplot(iw,jw,i);
m_proj('stereographic','lat',90,'long', 300,'radius',35,'rect','on')
m_grid('linewi',1,'tickdir','out',...
'xtick',[],'ytick',[])
m_quiver(long(1:10:361,1:10:361), lat(1:10:361,1:10:361),-C1(1:10:361,1:10:361),-C2(1:10:361,1:10:361),2,'k')
m_coast('patch',[.6 .6 .6],'edgecolor','k')
axis on
set(gca,'Xticklabel',[])
set(gca,'Yticklabel',[])
title(strcat('EOF:',num2str(i),'/',num2str(expv(i)),'%'));
else
subplot(iw,jw,iw*jw-1);
plot(pc_u(:,1),'b','DisplayName','PC of u');
hold on
plot(pc_v(:,1),'r','DisplayName','PC of v');
grid on
xlabel('time')
title('PC')
legend(num2str([1:N1]'),2);
box on
subplot(iw,jw,iw*jw);
plot(pc_u(:,2),'b','DisplayName','PC of u');
hold on
plot(pc_v(:,2),'r','DisplayName','PC of v');
grid on
xlabel('time')
title('PC')
legend(num2str([1:N1]'),2);
box on
end
end
Sophia
Sophia 2017-5-16

0 个投票

the total again is 99.9%
clear all; clc;
data= load('north_x_y_lat_lon');
datacoord = reshape(data, 361,361,4);
lat = squeeze(datacoord(:,:,3));
long = squeeze(datacoord(:,:,4));
rlong = long*pi/180.;
rlat = lat*pi/180.;
load nsidc_sid_1979_2014.mat
N1=3;
[eof_maps,pc,expv] = eof(t_u,3);
[eof_maps1,pc1,expv1] = eof(t_v,3);
figure(1);clf;iw=1;jw=N1+3;
set(gcf,'MenuBar','none');
for i=1:iw*jw
if i<= iw*jw-3
C1 = squeeze(eof_maps(:,:,i));
C2 = squeeze(eof_maps1(:,:,i));
pc_u = pc';
pc_v = pc1';
subplot(iw,jw,i);
m_proj('stereographic','lat',90,'long', 300,'radius',35,'rect','on')
m_grid('linewi',1,'tickdir','out',...
'xtick',[],'ytick',[])
m_quiver(long(1:10:361,1:10:361), lat(1:10:361,1:10:361),-C1(1:10:361,1:10:361),-C2(1:10:361,1:10:361),2,'k')
m_coast('patch',[.6 .6 .6],'edgecolor','k')
axis on
set(gca,'Xticklabel',[])
set(gca,'Yticklabel',[])
title(strcat('EOF:',num2str(i),'/',num2str(expv(i)),'%'));
else
subplot(iw,jw,iw*jw-2);
plot(pc_u(:,1),'b','DisplayName','PC of u');
hold on
plot(pc_v(:,1),'r','DisplayName','PC of v');
grid on
xlabel('time')
title('PC')
legend(num2str([1:N1]'),2);
box on
subplot(iw,jw,iw*jw-1);
plot(pc_u(:,2),'b','DisplayName','PC of u');
hold on
plot(pc_v(:,2),'r','DisplayName','PC of v');
grid on
xlabel('time')
title('PC')
legend(num2str([1:N1]'),2);
box on
subplot(iw,jw,iw*jw);
plot(pc_u(:,3),'b','DisplayName','PC of u');
hold on
plot(pc_v(:,3),'r','DisplayName','PC of v');
grid on
xlabel('time')
title('PC')
legend(num2str([1:N1]'),2);
box on
end
end

类别

在 帮助中心 和 File Exchange 中查找有关 Dimensionality Reduction and Feature Extraction 的更多信息

标签

提问:

Sophia
Sophia
2017-3-13

评论：

Sophia
Sophia
2017-5-16

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by