Finding max peak between two datetime values

Hi

I have a cluster of peaks that I want to make into one, but this is an issue. I have a function crossing_V7 that identifies when a peak starts (goes crosses the limit Abonnemang) and ends. Now I want a vector of the same length that has the height of every max peak in every cluster.

clear all
clc

load EFFEKT %EFFEKT contains a value for every our of the year
Abonnemang=21200;

dt = datetime(2021,1,1:1:365);
et = hours(0:23)';
t = repmat(dt,length(et),1) + repmat(et,1,length(dt));
t=t(:);
%(1:744) means I only look at January to simplify
[t0_pos1,s0_pos1,t0_neg1,s0_neg1]= crossing_V7(EFFEKT(1:744),t(1:744),Abonnemang,'linear'); % positive (pos) and negative (neg) slope crossing points

t0_pos1(10)
t0_neg1(10)

Answer 1

[Mathieu NOE]

hello

try this

clear all
clc
load EFFEKT %EFFEKT contains a value for every our of the year
Abonnemang=21200;
dt = datetime(2021,1,1:1:365);
et = hours(0:23)';
t = repmat(dt,length(et),1) + repmat(et,1,length(dt));
t=t(:);
%(1:744) means I only look at January to simplify
[t0_pos1,s0_pos1,t0_neg1,s0_neg1]= crossing_V7(EFFEKT(1:744),t(1:744),Abonnemang,'linear'); % positive (pos) and negative (neg) slope crossing points
t0_pos1(10)
t0_neg1(10)

% added code 

for ci = 1:numel(t0_pos1)
    idx = (t>=t0_pos1(ci)&t<=t0_neg1(ci));
    [max_of_cluster(ci),ix] = max(EFFEKT(idx));
    tmp = t(idx);
    tmax(ci) = tmp(ix);
    
    %plot
    figure(ci)
    ind = find(idx);
    a = 2; % add some pre and post samples to better visualize the raw data curve
    ind = ind(1)-a:ind(end)+a;
    plot(t(ind),EFFEKT(ind),t0_pos1(ci),Abonnemang,'dr',t0_neg1(ci),Abonnemang,'dk',tmax(ci),max_of_cluster(ci),'dm');
    title(['Cluster # ' num2str(ci)]);
end

Answer 2

[Antoni Garcia-Herreros]

Hello Joel,

You could try something like this:

clear all
clc
load EFFEKT %EFFEKT contains a value for every our of the year
Abonnemang=21200;
dt = datetime(2021,1,1:1:365);
et = hours(0:23)';
t = repmat(dt,length(et),1) + repmat(et,1,length(dt));
t=t(:);
%(1:744) means I only look at January to simplify
[t0_pos1,s0_pos1,t0_neg1,s0_neg1]= crossing_V7(EFFEKT(1:744),t(1:744),Abonnemang,'linear'); % positive (pos) and negative (neg) slope crossing points
t0_pos1(10)

ans = datetime
   27-Jan-2021 05:24:37

t0_neg1(10)

ans = datetime
   27-Jan-2021 19:37:10

MAXVEC=zeros(size(s0_neg1)); % Array to store the maximum values for each segment
POSMAX(size(MAXVEC,2))=datetime; % Variable to store the position of the maximum for each segment

for i=1:size(s0_neg1,2) % Loop through the segments
    jmin=find(t>t0_pos1(i),1); % Find the points inside the segment 
    jmax=find(t>t0_neg1(i),1)-1;
    tj=[t0_pos1(i);t(jmin:jmax);t0_neg1(i)]; % tj are the timepoints conforming the segment
    sj=[s0_pos1(i);EFFEKT(jmin:jmax);s0_neg1(i)];% sj are the y-values conforming the segment
    [MAXVEC(i),pos]=max(sj);
    POSMAX(i)=tj(pos);
end
%% Plot
imax=744;
plot(t(1:imax),EFFEKT(1:imax))
hold on
plot(t(1:imax),Abonnemang*ones(1,imax))
plot(t0_pos1,s0_pos1,'r*')
plot(t0_neg1,s0_neg1,'r*')
plot(POSMAX,MAXVEC,'g*','MarkerSize',10)
ylim([0.95*Abonnemang max(EFFEKT(1:imax))])
xlim([t0_pos1(1) t0_neg1(end) ])
l=legend({'EFFEKT','Abonnemang','','','Maximum'});
l.Position = [0.37, 0.64, 0.22, 0.12];

Hope this helps!