i have this problem, what i suppose to do?
2 ビュー (過去 30 日間)
古いコメントを表示
the problem was like this.
>> DBFBA(model,3,Ex_Succ(e),Ex_glc(e),10)
Undefined function or variable 'e'.
here is my coding
function [DBFBASol] = DBFBA(model,list,targetRxn,substrateRxn,MaxKOs)
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
imax =10; % number of iterations (e.g. 1000-5000)
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%Function Parameters%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
dim = length(list);
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%The Bees Algorithm%%%%%%%%%%%%%%%%%%%%%%%%%%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%Parameters%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
n = round(dim + (dim/4)); % Number of scout bees (e.g. 40-100)
pop = n;
m = 15; % Number of selected locations (e.g. 10-50)
%e = 3; % Number of elite selected points (e.g. 10-50)
e = MaxKOs;
%nsp = 8; % Number of Bees around each selected locations ( except the ellit location )
nsp = round(m/2);
%nep = 6; % Number of Bees around each elite locations
nep = (e * 2);
ngh = 30; % Patch radius for neighbourhood search
sc = 1; % Shrinking constant; defined as percentage (%) and range is between 0-100
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
global HTABLE
HTABLE = java.util.Hashtable;
global MaxKnockOuts
%%%%%%%%%%%%%%%%%
%Reactions/Genes%
%%%%%%%%%%%%%%%%%
rxnok = 1;
geneok = 1;
for i = 1:dim
if(~ ismember(list{i}, model.rxns)), rxnok = 0; end
if(~ ismember(list{i}, model.genes)),geneok = 0; end
end
if geneok
display('assuming list is genes');
elseif rxnok
display('assuming list is reactions');
else
display('list appears to be neither genes nor reactions: aborting');
return;
end
if nargin < 5
MaxKnockOuts = (dim * 0.1);
else
MaxKnockOuts = MaxKOs;
end
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
format long
aa = double(zeros(1,pop));
fit = double(zeros(1,n));
nghx=ngh; %%define patch size
tstart = clock;
%%%%%%%%Initial Random generation %%%%%%
for(i=1:n)
perm= randperm(dim);
for k =1:MaxKnockOuts
bPos(perm(k),i)=1;
end
end
for(i=1:n)
rxn_vector = bPos(:,i);
rxnList = list(logical(rxn_vector));
nummutations = sum(rxn_vector);
if nummutations > MaxKnockOuts
continue;
end
[isValidRxn,removeInd] = ismember(rxnList,model.rxns);
removeInd = removeInd(isValidRxn);
modelKO = model;
modelKO.ub(removeInd) = 0;
modelKO.lb(removeInd) = 0;
FBAsolution = optimizeCbModel(modelKO);
growthrate = FBAsolution.f;
if growthrate < .10
continue;
end
modelFixGR = modelKO;
gamma = 0.9*growthrate;
new_gr = FBAsolution.f - gamma;
modelFixGR = changeRxnBounds(modelKO,modelKO.rxns(modelKO.c==1),new_gr,'l');
modelFixGR = changeObjective(modelFixGR,targetRxn);
solMin = optimizeCbModel(modelFixGR,'max');
minProd = solMin.f;
fit(i) = minProd * (.98^nummutations)* growthrate;
end
%%%%%%%%%%%%End of initial random generation
%%%%%%%Run until maximum number of iteration met%%%%%%%%%
for iter = 1 : imax
disp(sprintf('Interation Number: %02.0f',iter));
%%%%Sorting fitnesses & positions
[fit sortedFit bPosSort]= funcSort(fit, bPos, dim, n);
%%%%%%%%%%Choosing best m
for(j= 1 : m)
for(d= 1 : dim)
candidx(d,j) = double(bPosSort(d,j));
end
end
%%%%%%Search inrando the neighbourhood
pop = candidx(:,1:m);
popold = zeros(size(pop)); % toggle population
val = zeros(1,m); % create and reset the "cost array"
bestmem = zeros(1,dim); % best population member ever
bestmemit = zeros(1,dim); % best population member in iteration
nfeval = 0; % number of function evaluations
CR = 0.5;
F = 1;
refresh = 5;
refresh = floor(refresh);
%------Evaluate the best member after initialization----------------------
ibest = 1; % start with first population member
for i=1 : m
rxn_vector = pop(:,i);
rxnList = list(logical(rxn_vector));
nummutations = sum(rxn_vector);
if nummutations > MaxKnockOuts
continue;
end
[isValidRxn,removeInd] = ismember(rxnList,model.rxns);
removeInd = removeInd(isValidRxn);
modelKO = model;
modelKO.ub(removeInd) = 0;
modelKO.lb(removeInd) = 0;
FBAsolution = optimizeCbModel(modelKO);
growthrate2 = FBAsolution.f;
if growthrate2 < .10
continue;
end
modelFixGR = modelKO;
gamma = 0.9*growthrate;
new_gr = FBAsolution.f - gamma;
modelFixGR = changeRxnBounds(modelKO,modelKO.rxns(modelKO.c==1),new_gr,'l');
modelFixGR = changeObjective(modelFixGR,targetRxn);
solMin = optimizeCbModel(modelFixGR,'max');
minProd = solMin.f;
val(i) = -minProd * (.98^nummutations)* growthrate;
end
bestval = val(1); % best objective function value so far
nfeval = nfeval + 1;
if (val(i) < bestval) % if member is better
ibest = i; % save its location
bestval = val(i);
end
bestmemit = pop(: , ibest); % best member of current iteration
bestvalit = bestval; % best value of current iteration
bestmem = bestmemit; % best member ever
%------DE-Minimization---------------------------------------------
%------popold is the population which has to compete. It is--------
%------static through one iteration. pop is the newly--------------
%------emerging population.----------------------------------------
pm1 = zeros(dim, m); % initialize population matrix 1
pm2 = zeros(dim, m); % initialize population matrix 2
pm3 = zeros(dim, m); % initialize population matrix 3
pm4 = zeros(dim, m); % initialize population matrix 4
pm5 = zeros(dim, m); % initialize population matrix 5
bm = zeros(dim, m); % initialize bestmember matrix
ui = zeros(dim, m); % intermediate population of perturbed vectors
mui = zeros(dim, m); % mask for intermediate population
mpo = zeros(dim, m); % mask for old population
rot = (0:1:m-1); % rotating index array (size NP)
rotd= (0:1:dim-1); % rotating index array (size D)
rt = zeros(m); % another rotating index array
rtd = zeros(dim); % rotating index array for exponential crossover
a1 = zeros(m); % index array
a2 = zeros(m); % index array
a3 = zeros(m); % index array
a4 = zeros(m); % index array
a5 = zeros(m); % index array
ind = zeros(4);
popold = pop; % save the old population
ind = randperm(4); % index pointer array
a1 = randperm(m); % shuffle locations of vectors
rt = rem(rot+ind(1),m); % rotate indices by ind(1) positions
a2 = a1(rt+1); % rotate vector locations
rt = rem(rot+ind(2),m);
a3 = a2(rt+1);
rt = rem(rot+ind(3),m);
a4 = a3(rt+1);
rt = rem(rot+ind(4),m);
a5 = a4(rt+1);
pm1 = popold(:,a1); % shuffled population 1
pm2 = popold(:, a2); % shuffled population 2
pm3 = popold(:, a3); % shuffled population 3
pm4 = popold(:, a4); % shuffled population 4
pm5 = popold(:, a5); % shuffled population 5
for i=1:m % population filled with the best member
bm(:,i) = bestmemit; % of the last iteration
end
mui = rand(dim, m) < CR; % all random numbers < CR are 1, 0 otherwise
ui = pm3 + F*(pm1 - pm2); % differential variation
ui = popold.*mpo + (ui.*mui).^2; % crossover
%-----Select which vectors are allowed to enter the new population------------
for i=1:m
rxn_vector = ui(:,i);
rxnList = list(logical(rxn_vector));
nummutations = sum(rxn_vector);
if nummutations > MaxKnockOuts
continue;
end
[isValidRxn,removeInd] = ismember(rxnList,model.rxns);
removeInd = removeInd(isValidRxn);
modelKO = model;
modelKO.ub(removeInd) = 0;
modelKO.lb(removeInd) = 0;
FBAsolution = optimizeCbModel(modelKO);
growthrate3 = FBAsolution.f;
if growthrate3 < .10
continue;
end
modelFixGR = modelKO;
gamma = 0.9*growthrate;
new_gr = FBAsolution.f - gamma;
modelFixGR = changeRxnBounds(modelKO,modelKO.rxns(modelKO.c==1),new_gr,'l');
modelFixGR = changeObjective(modelFixGR,targetRxn);
solMin = optimizeCbModel(modelFixGR,'max');
minProd = solMin.f;
tempval(i) = -minProd * (.98^nummutations)* growthrate;
end
temp = tempval(1,:);
for i=1:temp
nfeval = nfeval + 1;
if (tempval(i) <= val(i)) % if competitor is better than value in "cost array"
pop(:, i) = ui(:,i); % replace old vector with new one (for new iteration)
val(i) = tempval(i); % save value in "cost array"
%----we update bestval only in case of success to save time-----------
if (min(tempval) < bestval) % if competitor better than the best one ever
bestval = min(tempval); % new best value
bestmem = ui(:,i); % new best parameter vector ever
end
end
end %---end for imember=1:NP
bestmemit = bestmem; % freeze the best member of this iteration for the coming
% iteration. This is needed for some of the strategies.
%----Output section----------------------------------------------------------
if (refresh > 0)
if (rem(iter,refresh) == 0)
% fprintf(1,'Iteration: %d, Best: %+5.2d, F: %f, CR: %f, NP: %d\n',iter,bestval,F,CR,m);
% for n=1:m
% fprintf(1,'best(%d) = %f\n',m,bestmem(m));
% end
end
end
if (val(1,m) < sortedFit(1,m))
bPosSort(:,m) = ui(:,m);
sortedFit(1,m) = val(1,m);
end
%%%end of Neighbourhood Search
%%%Shrink all the patches using the shrinking constant (sc) variable
nghx=nghx*((100-sc)/100);
%%%Send rest of the bees for random search...
%%%%%Number of bees for random search
ranSearchBees= n - m;
for(k= 1 : ranSearchBees)
perm= randperm(dim);
for e =1:MaxKnockOuts
bPosSortcand(perm(e),k)=1;
end
end
for(k= 1 : ranSearchBees)
rxn_vectorSortcand = bPosSortcand(:,k);
rxnListSortcand = list(logical(rxn_vectorSortcand));
nummutationsSortcand = sum(rxn_vectorSortcand);
if nummutationsSortcand > MaxKnockOuts
continue;
end
[isValidRxnSortcand,removeIndSortcand] = ismember(rxnListSortcand,model.rxns);
removeIndSortcand = removeIndSortcand(isValidRxnSortcand);
modelKOSortcand = model;
modelKOSortcand.ub(removeIndSortcand) = 0;
modelKOSortcand.lb(removeIndSortcand) = 0;
FBAsolutionSortcand = optimizeCbModel(modelKOSortcand);
growthrateSortcand = FBAsolutionSortcand.f;
if growthrateSortcand < .10
continue;
end
modelFixGRSortcand = modelKOSortcand;
gamma = 0.9*growthrate;
new_grSortcand = FBAsolutionSortcand.f - gamma;
modelFixGRSortcand = changeRxnBounds(modelKOSortcand,modelKOSortcand.rxns(modelKOSortcand.c==1),new_grSortcand,'l');
modelFixGRSortcand = changeObjective(modelFixGRSortcand,targetRxn);
solMinSortcand = optimizeCbModel(modelFixGRSortcand,'max');
minProdSortcand = solMinSortcand.f;
sortedFitcand(k) = -minProdSortcand * (.98^nummutations)* growthrate;
if (sortedFitcand(1,k) < sortedFit(1,k+m))
bPosSort(:,m+k) = bPosSortcand(:,k);
sortedFit(1,k+m) = sortedFitcand(1,k);
end
end
[fit sortedFit bPosSort]= funcSort(sortedFit, bPosSort, dim, n);
template(iter,:) = double(sortedFit(1,:));
sortedFit(1,1);
v(iter,:) = sortedFit(1,1);
end
minProdFixGR = sortedFit(1,1);
Scores = sortedFit;
BPCY = sortedFit(1,1);
t = bPosSort(:,1);
r=list(logical(t));
Best = t;
DBFBASol = GetDBFBASol(model, list, targetRxn,substrateRxn, bPos,Best, Scores,BPCY,rxnok);
sort_temp = sort(template);
Final_Result = v(1:iter,1);
plot(Final_Result)
title('Differential Bees Flux Balance Analysis')
xlabel('Generation Number')
ylabel('Objective Function')
saveas(gcf, '/Users/fahmiarieef/desktop/psm/references/info/result/Objective_Function', 'jpg')
[biomassValues,targetValues,lineHandle] = productionEnvelope(model,r,'k',targetRxn,model.rxns(model.c==1));
saveas(gcf, '/Users/fahmiarieef/desktop/psm/references/info/result/Production_Envelope', 'jpg')
End_Time = etime(clock,tstart)
end
can anyone tell me which part i did wrong?
0 件のコメント
回答 (1 件)
Matt J
2015 年 12 月 13 日
編集済み: Matt J
2015 年 12 月 13 日
The problem is at the command line call, not inside DFBA()
>> DBFBA(model,3,Ex_Succ(e),Ex_glc(e),10)
Undefined function or variable 'e'.
The variable "e" does not exist.
4 件のコメント
Matt J
2015 年 12 月 13 日
編集済み: Matt J
2015 年 12 月 13 日
thank you for replying. but i already initialize variable e at the beginning.... e = MaxKOs
That's an excerpt from inside your function DBFBA, but you are also referring to "e" outside your function, at the command line.
>> DBFBA(model,3,Ex_Succ(e),Ex_glc(e),10)
The command line has its own workspace where "e" is something you haven't defined.
John D'Errico
2015 年 12 月 13 日
Before you call a function, NOTHING that is created inside the function exists at that point in time. So you are trying to use e as a variable, before e is ever defined.
参考
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!