Why does a for loop that has column vectors as input return row vectors as output

% Water System Blending Model
% Written by
% Data for Well Pumpage and Metered Usage was provided by the Security Public Water System (PWS)
% Columns cited in notes are from the Model tab in Security_Blending_Data.xlsx
 
[td]=xlsread('Security_Blending_Data.xlsx',1,"A3:A53");     %year
[Z1PumpTot]=xlsread('Security_Blending_Data.xlsx',1,"B3:B53");   %The "1" indicates the 1st excel tab shoudl be used in the work book
[Z1MassTot]=xlsread('Security_Blending_Data.xlsx',1,"C3:C53");
[Z3PumpTot]=xlsread('Security_Blending_Data.xlsx',1,"D3:D53");
[Z3MassTot]=xlsread('Security_Blending_Data.xlsx',1,"E3:E53");
[SorW]=xlsread('Security_Blending_Data.xlsx',1,"F3:F53");       %Column is 0 or 1 to designate summer or winter
 
td = zeros(1,51);                   %Preallocation of voctor size to increase speed of computation
 
 
for t=1:length(td)             %Loop to determine the Avg concnetration of PFHxS in Drinking WAter for each zone over time 
  
Z1Conc(t)=Z1MassTot(t)/Z1PumpTot(t);    %Total Mass PFHxS in Z1(Column 40)
Z3Conc(t)=Z3MassTot(t)/Z3PumpTot(t);    %Total Mass PFHxS in Z3(Column 52)
TotPS(t)=Z1PumpTot(t)+Z3PumpTot(t);     %Total Pumped Supply (Column 57)
 
    if SorW(t)==0                   %Z1 Percent Usage of Total Pumped Supply (Column 63)
        Z1PerTotPS(t)=0.297430058;
        
    else Z1PerTotPS(t)=0.361336585;
    end
    
    if SorW(t)==0                   %Z2 Percent Usage of Total Pumped Supply (Column 64)
        Z2PerTotPS(t)=0.367893872;
        
    else Z2PerTotPS(t)=0.359354700;
        
    end
    
    if SorW(t)==0                   %Z3 Percent Usage of Total Pumped Supply (Column 65)
        Z3PerTotPS(t)=0.334676069;
        
    else Z3PerTotPS(t)=0.279308716;
    end
        
 
Z1F(t)=Z1PerTotPS(t)*TotPS(t);             %Z1 Usage (based on % usage of TPS)(Column 66)
Z2F(t)=Z2PerTotPS(t)*TotPS(t);             %Z2 Usage (based on % usage of TPS)(Column 67)
Z3F(t)=Z3PerTotPS(t)*TotPS(t);             %Z3 Usage (based on % usage of TPS)(Column 68)
 
 
    if Z1PumpTot(t)>Z1F(t)                 %Z1 to Z2 Flow (Column 69)
        Z1toZ2F(t)=Z1PumpTot(t)-Z1F(t);
        
    else Z1toZ2F(t)=0;
    end
 
    if Z3PumpTot(t)>Z3F(t)                 %Z3 to Z2 Flow (Column 72)
        Z3toZ2F(t)=Z3PumpTot(t)-Z3F(t);
        
    else Z3toZ2F(t)=0;
    end
 
    if Z1PumpTot(t)<Z1F(t)                 %Z2 to Z1 Flow (Column 70)         
        Z2toZ1F(t)=Z3toZ2F(t)-Z2F(t);
        
    else Z2toZ1F(t)=0;
    end 
    
    if Z3PumpTot(t)<Z3F(t)                 %Z2 to Z3 Flow (Column 71)             
        Z2toZ3F(t)=Z1toZ2F(t)-Z2F(t);
        
    else Z2toZ3F(t)=0;
    end    
  
    if Z1toZ2F(t)>0                        %Z1 to Z2 Mass Load (Column 73)
        Z1toZ2M(t)=Z1toZ2F(t)*Z1Conc(t);
        
    else Z1toZ2M(t)=0;
    end
    
    if Z3toZ2F(t)>0                        %Z3 to Z2 Mass Load (Column 74)
        Z3toZ2M(t)=Z3toZ2F(t)*Z3Conc(t);
        
    else Z3toZ2M(t)=0;
    end
    
Z2TotM(t)=(Z1toZ2M(t)+Z3toZ2M(t))/(Z1toZ2F(t)+Z3toZ2F(t));   % Z2 Total Mass of PFHxS (Column 75)
 
    if Z2TotM(t)>0                         %Z2 to Z1 Mass Load (Column 76)
         Z2toZ1M(t)=Z2TotM(t)*Z2toZ1F(t);
         
    else Z2toZ1M(t)=0;
    end
 
    if Z2toZ3F(t)>0                       %Z2 to Z3 Mass Load (Column 77)
         Z2toZ3M(t)=Z2toZ3F(t)*Z2TotM(t);
         
    else Z2toZ3M(t)=0;
    end
    
Z1PFHxS(t)=(Z2toZ1M(t)+Z1MassTot(t))/(Z1PumpTot(t)+Z2toZ1F(t));      %Avg Concentration of PFHxS in Z1 (Column 78)
Z2PFHxS(t)=(Z1toZ2M(t)+Z3toZ2M(t))/(Z1toZ2F(t)+Z3toZ2F(t));          %Avg Concentration of PFHxS in Z2 (Column 79)
Z3PFHxS(t)=(Z2toZ3M(t)+Z3MassTot(t))/(Z2toZ3F(t)+Z3PumpTot(t));       %Avg Concentration of PFHxS in Z3 (Column 79)
 
end

SO, my question is, that I load column vectors from the excel sheet, and my loop returns row vectors....why? Does this matter?

[Adam]

Well, it's pretty difficult to scan through all that and understand it, but you have this line:

td = zeros(1,51);                   %Preallocation of voctor size to increase speed of computation

which is clearly declaring a row vector, although I can't see on scanning through where that is actually used. In fact you appear to just overwrite the data you loaded with zeros and then use the size of that now zero array for your for loop condition and then never use the array again! So already there is some questionable code there even though it seems it is not specifically the cause of your results changing their shape.

What you don't seem to do is presize all the other arrays you create though so if they just grow in the loop they will by default be row vectors.

If you step through your code a line at a time though and look at the variables on the command line it should become pretty obvious when different variables are different shapes.

Whether it matters or not depends entirely on what you are using them for, though personally like you I do not like sections of code that start off with column vectors and end with row vectors, or vice versa, as it is asking for trouble!

[Wesser]

Thanks!

I wasn't aware of the following:

"What you don't seem to do is presize all the other arrays you create though so if they just grow in the loop they will by default be row vectors."

I converted the code to read the following and it works great! Thank you

[td]=xlsread('Security_Blending_Data.xlsx',1,"A3:A53");

[Z1PumpTot]=xlsread('Security_Blending_Data.xlsx',1,"B3:B53");

[Z1MassTot]=xlsread('Security_Blending_Data.xlsx',1,"C3:C53");

[Z3PumpTot]=xlsread('Security_Blending_Data.xlsx',1,"D3:D53");

[Z3MassTot]=xlsread('Security_Blending_Data.xlsx',1,"E3:E53");

[SorW]=xlsread('Security_Blending_Data.xlsx',1,"F3:F53");

[TimeXaxis]=xlsread('Security_Blending_Data.xlsx',1,"A3:A53");

Date=datetime(TimeXaxis,'ConvertFrom','excel');

%Preallocaiton of vector size to increase speed of computation

Z1Conc=zeros(51,1);

Z3Conc=zeros(51,1);

TotPS=zeros(51,1);

Z1PerTotPS=zeros(51,1);

Z2PerTotPS=zeros(51,1);

Z3PerTotPS=zeros(51,1);

Z1F=zeros(51,1);

Z2F=zeros(51,1);

Z3F=zeros(51,1);

Z1toZ2F=zeros(51,1);

Z3toZ2F=zeros(51,1);

Z2toZ3F=zeros(51,1);

Z2toZ1F=zeros(51,1);

Z1toZ2M=zeros(51,1);

Z3toZ2M=zeros(51,1);

Z2TotM=zeros(51,1);

Z2toZ1M=zeros(51,1);

Z2toZ3M=zeros(51,1);

Z1PFHxS=zeros(51,1);

Z2PFHxS=zeros(51,1);

Z3PFHxS=zeros(51,1);

[Adam]

I often forget which way round is default for an array, but you can easily test it on command line (where I test many many things out to enhance my understanding!) by just typing something like e.g.

a(20) = 1;

so long as the variable 'a' does not already exist. This will then automatically create an array of length 20, which will, by default, be a row, which you can see easily in the workspace. This is effectively the same as growing in a loop as it has to create the preceeding 19 elements (as zeros) in order to do what you ask it and make the 20th element of an array which does not yet exist a 1.