Help fixing error code, please!

Question

Macie Smith on 15 Apr 2020

0
Link

Direct link to this question

https://ch.mathworks.com/matlabcentral/answers/518208-help-fixing-error-code-please

Commented: Macie Smith on 15 Apr 2020

Hello! Any and all help would be greatly appreciated!

I'm trying to run my code (below), but keep getting the following error code:

Indexin position 1 is invalid. Array indices must be positive integers or logical values.
Error in M_Model (line 56)
        RHS(:,pp)=((1-alpha)*Ft(xp(:,pp),tt+1))+
        (alpha*(1-Y)*Ft(xpp(:,pp),tt+1))+(alpha*Y*w*Ft(xppp(:,pp),tt+1))+(alpha*Y*(1-w)*Ft(xpppp(:,pp),tt+1));

So here is the code to my chop-it function, with the code I'm trying to run below that. Thank you VERY much in advance for any help/insights on how to fix this!!

%CHOPIT
function [state2]=chopit(state1,bottom,top)
%CHOPIT truncates state values at a low of xcrit and a high of Cap
%input is vector of state values, followed by single values for xcrit and
%cap
%outputs a vector of truncated state values
state2=(state1<=bottom).*bottom+(state1>bottom).*(state1<=top).*state1+(state1>top).*top;
end
%CODE
%specify constants
term    = 250; % terminal time
Cap     = 10;  % cap or max state
xcrit   = 3 ;  % critical value of x [range 1 to 10]
nchoice = 2 ;  % number of choices (varied cost/benefit of fight)
%initialize arrays and vectors used in the program
RHS = zeros(Cap,nchoice); % at each time step, expected fitness of each choice for individual of a given state
opt = zeros(Cap,term);    % optimal choice at each state at each time
Ft  = zeros(Cap,term);    % an array with max expected fitness at each state at each time
%x=1-10 (rows) at each time t=1-20 (columns)
%at each time step, for each state, this will be the max fitness over the
%two choices, which we will calculate below so for now, fill this array
%with zeros
x = linspace(1,Cap,Cap)'; %column vector of state values
%initialize x' and x" arrays with a row for each state and column for each
%patch
xp  = zeros(Cap,nchoice); % xprime for each state value, each patch
xpp = zeros(Cap,nchoice); % x double prime for each state value, each patch
%Specify patch parameter values
alpha  = [0.25 0.25]; % probability of encountering another male
beta   = [0.95 0.95]; % probability of surviving to next time-step
Y      = [0.5 0.5];   % probablity of fighting
lambda = [0.5 0.5];   % rate of decay in RHP if no fight
w      = [(1/(1+(exp(x)))) (1/(1+(exp(x))))]; % probability of winning a fight
c      = [3 3];       % cost of losing a fight
v      = [1 1];       % gain of winning a fight
%Specify values of xprime, x double prime, etc
for pp = 1:nchoice %loop through the choices
    xp(:,pp)   = chopit(x-lambda(pp),xcrit,Cap);
    xpp(:,pp)  = chopit(x-lambda(pp),xcrit,Cap);
    xppp(:,pp) = chopit(x+v(pp),xcrit,Cap); %#ok<AGROW>
    xppp(:,pp) = chopit(x-c(pp),xcrit,Cap);
end
%specify terminal fitness
Ft((1:xcrit),end)=0; % fitness=0 at or below xcrit
Ft((4:7),end) = 0.7; % fitness = 0.7 for sneak
Ft((8:Cap),end)=1 ;  % fitness=1 for territorial
%make a loop that begins at one step before terminal time and goes back to t=1
for tt= term-1:-1:1   % loop backwards through time
    for pp=1:nchoice; % loop through each patch
        %dynamic programming equation
        RHS(:,pp)=((1-alpha(pp))*Ft(xp(:,pp),tt+1))+ (alpha(pp)*(1-Y(pp))*Ft(xpp(:,pp),tt+1))+(alpha(pp)*Y(pp)*w(pp)*Ft(xppp(:,pp),tt+1))+(alpha(pp)*Y(pp)*(1-w(pp))*Ft(xpppp(:,pp),tt+1));
        RHS(1:xcrit,pp)=0;
    end
    %now find maximum RHS (i.e. max fitness & associated choice)
    %the fit and opt arrays will hold all model output
    %want max function to return a column vector with the max value of each
    %row of RHS. Recall RHS consists of expected fitness of each choice
    %(columns) for each state value (rows) & want each row max 
    [Ft(:,tt),opt(:,tt)]= max(RHS,[],2);
end

13 Comments
Show 11 older commentsHide 11 older comments

Walter Roberson on 15 Apr 2020

Ft(xp(:,pp),tt+1)) asks to recall xp(:,pp) and use it as the first index into the array Ft. Are you certain that xp(:,pp) will be strictly positive integers?

Macie Smith on 15 Apr 2020