for-loop returns vector of 1xi when all input data are vectors of ix1....why does it do this? Does it matter?

It doesn't seem to matter with the calculation of my for-loop, as the results are correct, but I am wondering why my for-loop automatically returns vectors with the dimensions of 1xi rather than ix1. Is this mathlab's default mode? I know you can force the vector dimensions to change with the follow type of equation,

as=As';

but I would like to understand how mathlab knows to move "right" along a 1xi vector and also move "down" along a ix1 vector within the same for-loop. This is what seems to be happening, and again the results are successful, but is this a "problem" that just happens to not be causing trouble? Thank you.

%ESCIMO energy balance method for snow melting.
%Data from the ... Switzerland
clear all

x=load('C:\Users\wesser\Desktop\P2\ESCIMO\ESCIMOdata.dat'); %Load the file with data

%Field Observed Parameters
td=x(:,4);              %hour of the day
T=x(:,5);                %Air Temperature (K)
Relhum=x(:,6);       %Relative Humidity (%)
Wind=x(:,7);          %Wind Speed (m/s)
P=x(:,8);                %Precipitation (mm/h)
GlobalRad=x(:,9);          %Global Radiation (W/m2)
Inlong=x(:,10);         %Incoming Longwave Radiation (W/m2)
Snow_obs=x(:,11);           %SWE Observed

%Parameters (to be CALIBRATED)
PM=importdata('C:\Users\wesser\Desktop\P2\ESCIMO\parameter.txt'); % model parameters input

Amin=PM(1,:);           %Minimum Albedo
Aadd=PM(2,:);           %Additive Albedo
DPp=PM(3,:);            %Decline parameter (Positive temperatures)
DPn=PM(4,:);            %Decline parameter (negative temperatures)
SS=PM(5,:);             %Significant snowfall (mm/h)
Tp=PM(6,:);             %Phase transition Temperature (K)
Se=PM(7,:);             %Snow Emissivity 
SHF=PM(8,:);            %Soil Heat Flux (W/m2)

%Calibrated parameter ranges ------TBD
%Amin - # to #         %Minimum Albedo
%Aadd - # to #          %Additive Albedo
%DPp - # to #            %Decline parameter (Positive temperatures)
%DPn - # to #            %Decline parameter (negative temperatures)
%SS - # to #             %Significant snowfall (mm/h)
%Tp - # to #             %Phase transition Temperature (K)
%Se - # to #             %Snow Emissivity 
%SHF - # to #            %Soil Heat Flux (W/m2)

%PARAMETERS (used default)
SHCW=4180.0;            %Specific heat capacity of water (J/(kgK))
SHCS=2100.0;            %Specific heat capacity of snow (J/(kgK))
MH=337500.0;             %Melting heat (J/kg)
SBC=0.0000000567;       %Stefan-Boltzmann-Constant (W/(m2K^4))
CSH=2835500.0;          %Condensation/Sublimation heat (J/kg)

%Preallocation 
Albedo(1)=0;
SnowAge(1)=0;
ModeledSWE(1)=0;
Melt(1)=0;
Sublim(1)=0;
VapPresS(1)=0;
Pr(1)=0; 
Ps(1)=0;
Z(1)=0;
SurfTemp(1)=0;
VapPresA(1)=0;
ShortRadBal(1)=0;
LongRadBal(1)=0;
SenHeatFlux(1)=0;
LatHeatFlux(1)=0;
AdvFluxR(1)=0;
AdvFluxS(1)=0;

%Precipitation Loop
for t=2 : length(P);             %time series loop
    if T(t)>=Tp;       %Precip that is rain       
       Pr(t)=P(t);
    else
       Pr(t)=0;
    end

      if T(t)<Tp;      %Precip that is snow
         Ps(t)=P(t);
      else
         Ps(t)=0;
      end

      if Ps(t)>SS;         %Snow age
          SnowAge(t)=0;
      else
          if ModeledSWE(t-1)>0;
              SnowAge(t)=SnowAge(t-1)+0.04167;
          else
              SnowAge(t)=0;
          end
      end

      if T(t)>273.16;
          Z(t)=DPp;
      else
          Z(t)=DPn;
      end
      if Ps(t)>SS;    %Albedo
          Albedo(t)=Amin+Aadd;
      else
          if SnowAge(t)>0;
              Albedo(t)=Amin+(Albedo(t-1)-Amin)*exp(Z(t)*0.04167);
          else   
             Albedo(t)= 0;
          end
      end

      SurfTemp(t)=min(T(t),273.16);

      VapPresA(t)=6.1078*exp((17.08085*(T(t)-273.16))/(234.175+(T(t)-273.16)))*(Relhum(t)/100);

      if ModeledSWE(t-1)>0;
          VapPresS(t)=6.1078*exp((17.088085*(SurfTemp(t)-273.16))/(234.175+(SurfTemp(t)-273.16)));
      else
          VapPresS(t)=0;
      end

      ShortRadBal(t)=(1- Albedo(t))*GlobalRad(t);
      LongRadBal(t)=Inlong(t)-(Se*SBC*SurfTemp(t)^4);
      SenHeatFlux(t)=18.85*(0.18+0.098*Wind(t))*(T(t)-SurfTemp(t));
      LatHeatFlux(t)=32.82*(0.18+0.098*Wind(t))*(VapPresA(t)-VapPresS(t));

      if Pr(t)>0;
          AdvFluxR(t)=SHCW*(T(t)-273.16)*Pr(t)/3600;
      else
          AdvFluxR(t)=0;
      end

      if Ps(t)>0;
          AdvFluxS(t)=SHCS*(T(t)-273.16)*Ps(t)/3600;
      else
          AdvFluxS(t)=0;
      end

      EB(t)=ShortRadBal(t)+LongRadBal(t)+SenHeatFlux(t)+LatHeatFlux(t)+AdvFluxR(t)+AdvFluxS(t)+SHF;

      if ModeledSWE(t-1)>0;
         Sublim(t)=3600*LatHeatFlux(t)/CSH;
      else
         Sublim(t)=0;
      end

      if T(t)>273.16;
          if ModeledSWE(t-1)>0 && EB(t)>0;
             Melt(t)= min(EB(t)*3600/MH ,ModeledSWE(t-1));
          else
              Melt(t)=0;
          end
      else
          Melt(t)=0;
      end

      if ModeledSWE(t-1)>0;
          ModeledSWE(t)=max((ModeledSWE(t-1)+Pr(t)+Ps(t)+Sublim(t)-Melt(t)),0);
      else
          ModeledSWE(t)=max((Ps(t)+Sublim(t)-Melt(t)),0);
      end
  end

daynr=1:length(P);

dlmwrite('output.txt', transpose(ModeledSWE), 'newline','pc','delimiter', '\t') 
quit force

Answer 1

[Star Strider]

You can force a column vector by adding an extra dimension to the subscripts so that it saves by rows instead of by the default columns:

for k1 = 1:10
    row_vector(k1)  = k1^2;
    col_vector(k1,:) = k1^2;
end