How do I loop on multiple indicies?

I have the following code that performs the calculation, but I need to repeat that calculation for about 100 sets of input values. Specifically, I need to vary the values for paro, mmhgo, flowrate, sechii, and transit. I can use a for loop for one of the variables, but I do not know how to code for changing them all simultaneously. I imagine that the code could read one line (row) of input data from a file with each field (column) corresponding to one of my variables just mentioned. In other words, for one set of calculations the inputs would be paro=44.3, mmgho=0.022, flowrate=1.7e7, sechii=45, and transit=5. For another run, paro=49.27, mmhgo=0.09, flowrate=5.69e10, sechii=76, and transit=23. Any suggestions on how to do this are much appreciated.

 mixingconst=4; %number of times in one day that a day's volume of water is homogenized  
    sechii=76; %in cm
    flowrate=1.70E+10; %Flow rate in L day-1
    paro=49.29; %surface mean monthly PAR value mol m-2 day-1
    mmhgo=[0.09,0.022]; %initial MMHg concentration in ng L-1
    transit=14*mixingconst; %transit days * the number of mixing times each day
    kz=(1.44/(sechii/100))/10; %puts kz into units of 0.1 m-1
    kd=0.011; %mean value from incubation experiments
    depth=70; %this I hold constant
    z = 1:depth; %each layer is 0.1 m
        parz = paro*exp(-kz*z);  %light penetration equation
        layerconst=exp(-kd*parz); %the loss fraction at each layer
        mulconstant = sum(layerconst) / depth; %the loss fraction for the whole water column
        mmhgnew = mmhgo * cumprod(repmat(mulconstant, 1, transit)); %the progression of water column concentrations (cumulative product).
        mmhgmnew=mmhgnew*(flowrate/1e9); %the progression of water column mass (cumulative product)
        mmhgmo=mmhgo*(flowrate/1e9); %water column mass on day 0
        mmhgmloss=mmhgmo-mmhgmnew(transit) %cumulative mass lost

Answer 1

[Cam Salzberger]

Hello Peter,

The way you mentioned, having all the constants in a big matrix and looping through just the rows, would work fine. Or you can just have each constant as a vector of the same length, and loop through the indices. Something like this:

% Define parameters
paro = [44.3 49.27]; % Or however many entries you want
mmhgo = [0.022 0.09];
flowrate = [1.7e7 5.69e10];
sechii = [45 76];
transit = [5 23];

% Preallocate output variable(s)
mmhgmloss = zeros(size(paro));

% Perform calculation for each parameter set
for k = 1:numel(paro)
    kz=(1.44/(sechii(k)/100))/10;
    %... perform calculation indexing into each parameter ...
    mmhgmloss(k) = mmhgmo-mmhgmnew(transit(k));
end

Hope that helps give you a framework to get started.

-Cam

[Peter Weiss]

Thank you for the help. Can you tell me if I can call the matrix values from an imported matrix, or do I have to put the values in square brackets in the code?

[Walter Roberson]

Certainly you can import or calculate them.

paro = xlsread('paro.xlsx');

for example.

[Peter Weiss]

Thank you Walter. I assume that I can also output the values of mmhgmloss(k) to a matrix or table . I am looking for the command to do that.

[Cam Salzberger]

Well, the way it's set up in my code, it will be a matrix. In this case, I believed that you were only looping through a set of "paired" parameters, so there would only be a single array as output. If the value of mmhgmloss for each set of parameters is a vector though, then you would preallocate with something like:

mmhgmloss = zeros(size(paro,1),lengthOfMmhgmloss);

and you could assign each row with something like:

mmhgmloss(k,:) = mmhgmo-mmhgmnew(transit(k));

You could do a similar thing by preallocating the table and then assigning each value as you go. See the table function for creating them, and this post for a suggestion on prallocation.

-Cam