Hi,I tray RUN this code but i get tis error ( below the code)

% Please note that this section requires the toolbox m_map
% Now we would like to know the mean states and annual trends of MHW
% frequency, i.e. how many MHW events would be detected per year and how it
% changes with time.
[mean_freq,annual_freq,trend_freq,p_freq]=mean_and_trend_new(MHW,mhw_ts,1982,'Metric','Frequency');

Unrecognized function or variable 'MHW'.

% These four outputs separately represent the total mean, annual mean,
% annual trend and associated p value of frequency.
% This function could detect mean states and trends for six different
% variables (Frequency, mean intensity, max intensity, duration and total
% MHW/MCs days). 
metric_used={'Frequency','MeanInt','MaxInt','CumInt','Duration','Days'};
for i=1:6;
    eval(['[mean_' metric_used{i} ',annual_' metric_used{i} ',trend_' metric_used{i} ',p_' metric_used{i} ']=mean_and_trend_new(MHW,mhw_ts,1982,' '''' 'Metric' '''' ',' 'metric_used{i}' ');'])
end
% plot mean and trend
% It could be detected that, as a global hotspot, the oceanic region off
% eastern Tasmania exhibits significant positive trends of MHW metrics. 
figure('pos',[10 10 1500 1500]);
m_proj('mercator','lat',[-45 -37],'lon',[147 155]);
for i=1:6;
    subplot(2,6,i);
    eval(['mean_here=mean_' metric_used{i} ';']);
    eval(['t_here=trend_' metric_used{i} ';']);
    m_pcolor(lon_used,lat_used,mean_here');
    shading interp
    m_coast('patch',[0.7 0.7 0.7]);
    m_grid;
    colormap(jet);
    s=colorbar('location','southoutside');
    title(metric_used{i},'fontname','consolas','fontsize',12);
    
    subplot(2,6,i+6);
    eval(['mean_here=mean_' metric_used{i} ';']);
    eval(['t_here=trend_' metric_used{i} ';']);
    m_pcolor(lon_used,lat_used,t_here');
    shading interp
    m_coast('patch',[0.7 0.7 0.7]);
    m_grid;
    colormap(jet);
    s=colorbar('location','southoutside');
    title(['Trend-' metric_used{i}],'fontname','consolas','fontsize',12);
end
%% 5. Applying cluster algoirthm to MHW - A kmeans example.
% We get so many MHWs now.... Could we distinguish them into different
% gropus based on their metrics?
% Change it to matrix;
MHW_m=MHW{:,:};
% Extract mean, max, cumulative intensity and duration.
MHW_m=MHW_m(:,[3 4 5 7]); 
[data_for_k,mu,sd]=zscore(MHW_m);
% Determine suitable groups of kmeans cluster.
index_full=[];
cor_full=[];
for i=2:20;
    k=kmeans(data_for_k,i,'Distance','cityblock','maxiter',200);
    k_full=[];
    for j=1:i;
        k_full=[k_full;nanmean(data_for_k(k==j,:))];
    end
    
    k_cor=k_full(k,:);
    k_cor=k_cor(:);
    
    [c,p]=corr([data_for_k(:) k_cor]);
    
    index_full=[index_full;2];
    cor_full=[cor_full;c(1,2)];
    
    
end
figure('pos',[10 10 1500 1500]);
subplot(1,2,1);
plot(2:20,cor_full,'linewidth',2);
hold on
plot(9*ones(1000,1),linspace(0.6,1,1000),'r--');
xlabel('Number of Groups','fontsize',16,'fontweight','bold');
ylabel('Correlation','fontsize',16,'fontweight','bold');
title('Correlation','fontsize',16);
set(gca,'xtick',[5 9 10 15 20],'fontsize',16);
subplot(1,2,2);
plot(3:20,diff(cor_full),'linewidth',2);
hold on
plot(9*ones(1000,1),linspace(-0.02,0.14,1000),'r--');
xlabel('Number of Groups','fontsize',16,'fontweight','bold');
ylabel('First difference of Correlation','fontsize',16,'fontweight','bold');
title('First Difference of Correlation','fontsize',16);
set(gca,'fontsize',16);
% Use 9 groups.
k=kmeans(data_for_k,9,'Distance','cityblock','maxiter',200);
k_9=[];
prop_9=[];
for i=1:9;
    data_here=data_for_k(k==i,:);
    data_here=nanmean(data_here);
    data_here=data_here.*sd+mu;
    k_9=[k_9;data_here];
    prop_9=[prop_9;nansum(k==i)./size(data_for_k,1)];
end
loc_x=[1.5 1.5 1.5 2.5 2.5 2.5 3.5 3.5 3.5];
loc_y=[1.5 2.5 3.5 1.5 2.5 3.5 1.5 2.5 3.5];
text_used={['1: ' num2str(round(prop_9(1)*100)) '%'],['2: ' num2str(round(prop_9(2)*100)) '%'],['3: ' num2str(round(prop_9(3)*100)) '%'],...
    ['4: ' num2str(round(prop_9(4)*100)) '%'],['5: ' num2str(round(prop_9(5)*100)) '%'],['6: ' num2str(round(prop_9(6)*100)) '%'],...
    ['7: ' num2str(round(prop_9(7)*100)) '%'],['8: ' num2str(round(prop_9(8)*100)) '%'],['9: ' num2str(round(prop_9(9)*100)) '%']};
figure('pos',[10 10 1500 1500]);
h=subplot(2,2,1);
data_here=k_9(:,1);
data_here=reshape(data_here,3,3);
data_here(:,end+1)=data_here(:,end);
data_here(end+1,:)=data_here(end,:);
pcolor(1:4,1:4,data_here);
set(h,'ydir','reverse');
axis off
colormap(jet);
text(loc_x,loc_y,text_used,'fontsize',16,'horiz','center','fontweight','bold');
colorbar;
title('Durations','fontsize',16,'fontweight','bold');
h=subplot(2,2,2);
data_here=k_9(:,2);
data_here=reshape(data_here,3,3);
data_here(:,end+1)=data_here(:,end);
data_here(end+1,:)=data_here(end,:);
pcolor(1:4,1:4,data_here);
axis off
set(h,'ydir','reverse');
colormap(jet);
text(loc_x,loc_y,text_used,'fontsize',16,'horiz','center','fontweight','bold');
colorbar
title('MaxInt','fontsize',16,'fontweight','bold');
h=subplot(2,2,3);
data_here=k_9(:,3);
data_here=reshape(data_here,3,3);
data_here(:,end+1)=data_here(:,end);
data_here(end+1,:)=data_here(end,:);
pcolor(1:4,1:4,data_here);
axis off
set(h,'ydir','reverse');
colormap(jet);
text(loc_x,loc_y,text_used,'fontsize',16,'horiz','center','fontweight','bold');
colorbar
title('MeanInt','fontsize',16,'fontweight','bold');
h=subplot(2,2,4);
data_here=k_9(:,4);
data_here=reshape(data_here,3,3);
data_here(:,end+1)=data_here(:,end);
data_here(end+1,:)=data_here(end,:);
[x,y]=meshgrid(1:4,1:4);
pcolor(1:4,1:4,data_here);
set(h,'ydir','reverse');
axis off
colormap(jet);
text(loc_x,loc_y,text_used,'fontsize',16,'horiz','center','fontweight','bold');
colorbar
title('CumInt','fontsize',16,'fontweight','bold');
% Their associated SSTA patterns
% Calculate SSTA
time_used=datevec(datenum(1982,1,1):datenum(2016,12,31));
m_d_unique=unique(time_used(:,2:3),'rows');
ssta_full=NaN(size(sst_full));
for i=1:size(m_d_unique);
    date_here=m_d_unique(i,:);
    index_here=find(time_used(:,2)==date_here(1) & time_used(:,3)==date_here(2));
    ssta_full(:,:,index_here)=sst_full(:,:,index_here)-nanmean(sst_full(:,:,index_here),3);
end
sst_1993_2016=ssta_full(:,:,(datenum(1993,1,1):datenum(2016,12,31))-datenum(1982,1,1)+1);
time_used=MHW{:,:};
time_used=time_used(:,1:2);
start_full=datenum(num2str(time_used(:,1)),'yyyymmdd')-datenum(1993,1,1)+1;
end_full=datenum(num2str(time_used(:,2)),'yyyymmdd')-datenum(1993,1,1)+1;
for i=1:9;
    start_here=start_full(k==i);
    end_here=end_full(k==i);
    
    index_here=[];
    
    for j=1:length(start_here);
        period_here=start_here(j):end_here(j);
        index_here=[index_here;period_here(:)];
    end
    
    eval(['sst_' num2str(i) '=nanmean(sst_1993_2016(:,:,index_here),3);'])
    
end
color_used=hot;
color_used=color_used(end:-1:1,:);
figure('pos',[10 10 1500 1500]);
plot_index=[1 4 7 2 5 8 3 6 9];
for i=1:9;
    eval(['plot_here=sst_' num2str(i) ';']);
    subplot(3,3,plot_index(i));
    eval(['data_here=sst_' num2str(i) ';'])
    
    m_contourf(lon_used,lat_used,data_here',-3:0.1:3,'linestyle','none');
    
    
    if i~=3;
        m_grid('xtick',[],'ytick',[]);
    else;
        m_grid('linestyle','none');
    end
    m_gshhs_h('patch',[0 0 0]);
    colormap(color_used);
    caxis([0 2]);
    
    title(['Group (' num2str(i) '):' num2str(round(prop_9(i)*100)) '%'],'fontsize',16);
    
end
hp4=get(subplot(3,3,9),'Position'); 
s=colorbar('Position', [hp4(1)+hp4(3)+0.025 hp4(2) 0.025 0.85],'fontsize',14);
s.Label.String='^{o}C';

Index exceeds the number of array elements. Index must not exceed 12784.

Error in event_line (line 84)

y1=m90_plot(x1-datenum(data_start,1,1)+1);

^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

Error in an_example2 (line 53)

event_line(sst_full,MHW,mclim,m90,[1 2],1982,[2015 9 1],[2016 5 1]);

[Walter Roberson]

    eval(['[mean_' metric_used{i} ',annual_' metric_used{i} ',trend_' metric_used{i} ',p_' metric_used{i} ']=mean_and_trend_new(MHW,mhw_ts,1982,' '''' 'Metric' '''' ',' 'metric_used{i}' ');'])

I'm out. There are very very seldom good reasons to use eval()

Answer 1

[Torsten]

Variables and/or functions are missing to run your code.

According to the error message, m90_plot is an array of size 12784, but the number x1-datenum(data_start,1,1)+1 is bigger than 12784. Thus you get an access error.

Answer 2

[Image Analyst]

@B Your index is bigger than 12784, which is the max index m90_plot can take. Put these lines in and tell us what you see in the command window.

size(m90_plot)
fprintf('data_start = %g\n', data_start);
fprintf('datenum(data_start,1,1) = %g\n', datenum(data_start,1,1));
index = x1-datenum(data_start,1,1)+1
fprintf('x1-datenum(data_start,1,1)+1 = %g\n', index);

m90_plot is an array so the index must be an integer. Is index an integer or a floating point number.

datenum most likely returns a huge number. Look

datenum(2024, 10, 6)

ans = 739531

Not sure what you're attempting to do there but it doesn't seem right unless x1 is also a datenum value.

Also, don't use eval. Learn how to use fprintf

Instead of

eval(['mean_here=mean_' metric_used{i} ';']);
eval(['t_here=trend_' metric_used{i} ';']);

do

fprintf('mean_here = mean_%s\n', metric_used{i});
fprintf('t_here = trend_%s\n', metric_used{i});

See the FAQ: https://matlab.fandom.com/wiki/FAQ#Why_is_it_advised_to_avoid_using_the_%22eval%22_function?

Answer 3

[B]

Thank you

Still the same error

[Image Analyst]

Thank who? Who is your "Answer" supposed to be a comment to? I think you did not even see my Answer otherwise you would have answered when I said:

Put these lines in and tell us what you see in the command window.

size(m90_plot)
fprintf('data_start = %g\n', data_start);
fprintf('datenum(data_start,1,1) = %g\n', datenum(data_start,1,1));
index = x1-datenum(data_start,1,1)+1
fprintf('x1-datenum(data_start,1,1)+1 = %g\n', index);

So go back up and read my answer and put your reply in the comment below it.