Correlating values from a table using a for loop

I have a table that has hourly data (t) and by day (d) of chlorophyll concentrations. I am trying to find the maximum value of Chlorophyll of the surface and bottom values for each day as well as their corresponding depths and times. Does anyone know how to do this in a for loop??? I used varfun and @max in a table set up but now moving into larger data sets, a for loop would be a better route.

T1=table(d,t,Chl_Bottom, depth_chl_bottom, Chl_Surface, depth_chl_surface);

k=0;

for i = 1:length(d)

k=k+1;

smax= find(max(Chl_Surface(i))); %find the max of Chl_Surface by day

bmax = find(max(Chl_Bottom(i))); %find the max of Chl_Bottom by day

%find the time of the day when the peak is at its max

%find the depth of the day when the peak is at its max

end

[Sourabh Kondapaka]

Hi,

Could you upload the table 'T1'

[Jacqueline Chrabot]

See Attached.

[Sourabh Kondapaka]

As this seems to be your first question, please go through this tutorial on how to ask a question.

Answer 1

[Mathieu NOE]

hello

hope my code can further help you - you don't need any for loop to find the max value (and the corresponding time index) of a vector.

enjoy

T = readtable('data.csv')
C = table2cell(T);
% data organized as follows : Date,time,depth_chl_bottom,Chl_Bottom,depth_chl_surface,Chl_Surface

Date = C(:,1);
time = C(:,2);
depth_chl_bottom = cell2mat(C(:,3));
Chl_Bottom = cell2mat(C(:,4));
depth_chl_surface = cell2mat(C(:,5));
Chl_Surface = cell2mat(C(:,6));

% % T1=table(d,t,Chl_Bottom, depth_chl_bottom, Chl_Surface, depth_chl_surface);
%  k=0;
% for i = 1:length(Date)
%    k=k+1;
%         smax= find(max(Chl_Surface(i)));    %find the max of Chl_Surface by day
%         bmax = find(max(Chl_Bottom(i)));    %find the max of Chl_Bottom by day 
%                                                  %find the time of the day when the peak is at its max
%                                                 %find the depth of the day when the peak is at its max
% end

[Chl_Surface_max_value,time_index_max_Chl_Surface] = max(Chl_Surface);
[Chl_Bottom_max_value,time_index_max_Chl_Bottom] = max(Chl_Bottom);

x_axis = 1:length(Date);

figure(1),
subplot(2,1,1),plot(x_axis,Chl_Surface,'b',x_axis(time_index_max_Chl_Surface),Chl_Surface_max_value,'ro');
subplot(2,1,2),plot(x_axis,Chl_Bottom,'b',x_axis(time_index_max_Chl_Bottom),Chl_Bottom_max_value,'ro');

[Sourabh Kondapaka]

Hi Mathieu NOE,

When reading the data as a table we can directly work with the table instead of converting it to a different format.

Jacqueline Chrabot had asked for a max Chl_Surface and Chl_Bottom for each day. Your code will find max of the aforementioned values for the entire dataset.

Answer 2

[Sourabh Kondapaka]

Hi Jacqueline Chrabot,

The following code will help you in achieving what you want:

%Read 'Data.csv' as a table
T =readtable('DATA.csv');

% As you want max values of Chl_Surface and Chl_Bottom for each day, we will need
% all the unique dates in the table first
unique_dates = unique(T.Date);

% Iterating over each unique date and finding max Chl_Surface and Chl_Bottom for that day:
for i = 1 : length(unique_dates)
    
    %For the current unique day
    fprintf("For the day %s is \n", string(unique_dates(i)));
    
    % Extracting a part of T for the current date.
    t = T(T.Date == unique_dates(i), :);
    
    % Extracting max Chl_Surface value and the index where it occurs
    [maxSurfaceValue, maxSurfaceIdx] = max(t.Chl_Surface);
    
    % Display the max Chl_Surface value and the depth and time it occurs
    disp("Max Chl_Surface is :");
    disp(T(maxSurfaceIdx, :));
    
    % Extracting max Chl_Bottom value and the index where it occurs
    [maxBottomValue, maxBottomIdx] = max(t.Chl_Bottom); 
    
    % Display the max Chl_Bottom value and the depth and time it occurs
    disp("Max Chl_Bottom is : ");
    disp(T(maxBottomIdx, :));   
    
end

A similar question about extracting max value in a column has been asked here

Check the documention for "max()" function here

[Jacqueline Chrabot]

Thank you so much, this did exactly what I needed.

[Jacqueline Chrabot]

How would I change this code to find not the max but the first and last peak per day of the chlorophyll bottom and surface???

[Sourabh Kondapaka]

Hi Jacqueline Chrabot,

Can you briefly explain what you mean by first and last peak of chlorophyll_bottom for a day ?

Do you mean that there should be two peaks of the same value per day for chlorophyll_bottom ?

To get started with the fundamentals of matlab, I would recommend a free Matlab Onramp Course

[Jacqueline Chrabot]

There are multiple peaks during a day for each the chlorophyll surface reading and chlorophyll bottom reading. You originally helped provide me a script finding the mean of those peaks for the surface and bottom per day (and the corresponding depth and time associated with the mean) But what if I wanted just the first peak and the last peak of the surface and bottom per day. I'm thinking that the first peak would be more accurate of a reading rather than taking the mean ( as taking the mean would include maybe some noise).

[Jacqueline Chrabot]

So instead of using the max command is there another function that might give me the first or last peak per day. Giving me four outputs per day (two for surface chlorophyll surface and two for chlorophyll bottom)

Answer 3

[Peter Perkins]

Not sure why you think a "loop is better". Try this:

>> t = table([1;1;1;1;2;2;2;2;2;2],rand(10,1),rand(10,1),'VariableNames',["x" "y" "z"])
t =
  10×3 table
    x       y           z   
    _    ________    _______
    1     0.40286    0.67573
    1     0.69316     0.2173
    1     0.96076    0.91596
    1    0.039119    0.75431
    2     0.33803     0.5145
    2     0.99243    0.87343
    2     0.40484    0.90224
    2     0.65143    0.22551
    2     0.96835    0.27113
    2     0.74077    0.83571
>> rowfun(@myfun,t,'GroupingVariable','x','OutputVariableNames',["ymax" "zmax"])
ans =
  2×4 table
    x    GroupCount     ymax       zmax  
    _    __________    _______    _______
    1        4         0.98035    0.13753
    2        6         0.88784    0.45141

where

function [ymax,zmax] = myfun(y,z)
[ymax,imax] = max(y);
zmax = z(imax);
end

With your real data, same thing, just more verbose because if the long names:

t = readtable('DATA.csv');
rowfun(@myfun,t,'GroupingVariable','Date', ...
    'OutputVariableNames',["time_bottom_max" "depth_chl_bottom_max" "Chl_Bottom_max" "time_surface_max" "depth_chl_surface_max" "Chl_Surface_max"])

function [time_bottom_max,depth_chl_bottom_max,Chl_Bottom_max,time_surface_max,depth_chl_surface_max,Chl_Surface_max] = myfun(time,depth_chl_bottom,Chl_Bottom,depth_chl_surface,Chl_Surface)
    [Chl_Bottom_max,imax] = max(Chl_Bottom);
    time_bottom_max = time(imax);
    depth_chl_bottom_max = depth_chl_bottom(imax);
    
    [Chl_Surface_max,imax] = max(Chl_Surface);
    time_surface_max = time(imax);
    depth_chl_surface_max = depth_chl_surface(imax);
end