Write multiple variables from a function

I'm aware this is probably a horribly inefficient way of completing my work. What I'm trying to do is run through a workflow with a variety of different permutations of variables and get the large range of outputs for me to analyse.

What I've wrote up so far is:

% Define the CO2 Saturation sampling percentages (every 0.5%)
sat_samp_int=linspace(0,1,200);

% Generate the various saturation logs for the different scenarios
sw_sats=Sw*sat_samp_int;
co2_sats=CO2_Sat*sat_samp_int;

% Loop through the different saturations

for i = 1:200
    qq1=sw_sats(:,i)./k_brine;
    qq2=co2_sats(:,i)./Kb.';
end

In essence, I want to be able to generate the 200 different scenarios but I think the way I'm writing this so far is that each iteration is overwriting the previous loop so I'm only left with the output with reference to the final column of "sw_sats" and "co2_sats".

Is there a way to implement it so that I can actually be left with all my different iterations visible for me to view? Even if it leaves me with a massive array that is workable for me.

[James Tursa]

You probably can do all of this without a loop. What are the sizes of all of the variables involved?

[Martyn Steel]

Hi James,

sw_sats is a 7071x200 array

co2_sat is 7071x200 array

k_brine is a 1x1500 array

Kb is a 1x1500 array

[Star Strider]

sw_sats is a 7071x200 array

co2_sat is 7071x200 array

k_brine is a 1x1500 array

Kb is a 1x1500 array

That means that ‘Sw’ and ‘CO2_Sat’ are (7071 x 1) vectors. Make this straightforward with:

sat_samp_int=linspace(0,1,1500);

and they are now instead (7071 x 1500) matrices. With that, everything has compatible dimensions and straightforward matrix division is possible.

qq1=sw_sats./k_brine;

qq2=co2_sats./Kb;

Make appropriate transpose or reshape operations to get the result you want (since I have no idea what that is).

.

[Martyn Steel]

Hi Star Strider,

I fully think your logic of just making everything the same shape is actually just going to be my best approach moving forwards.

Thank you for the response!

[Star Strider]

My pleasure!

I’ll post it as an Answer now, as well.

Answer 1

[Star Strider]

sw_sats is a 7071x200 array

co2_sat is 7071x200 array

k_brine is a 1x1500 array

Kb is a 1x1500 array

That means that ‘Sw’ and ‘CO2_Sat’ are (7071 x 1) vectors. Make this straightforward with:

sat_samp_int=linspace(0,1,1500);

and they are now instead (7071 x 1500) matrices. With that, everything has compatible dimensions and straightforward matrix division is possible.

qq1=sw_sats./k_brine;

qq2=co2_sats./Kb;

Make appropriate transpose or reshape operations to get the result you want (since I have no idea what that is).

.

Answer 2

[James Tursa]

Maybe something like this does what you want, with each qq1 and qq2 2D page results in the first two dimensions.

[m n] = size(sw_sats);
qq1 = reshape(sw_sats,m,1,n) ./ k_brine;
qq2 = reshape(co2_sats,m,1,n) ./ Kb;

That is,

qq1(:,:,1) would be the result of sw_sats(:,1) ./ k_brine

qq1(:,:,2) would be the result of sw_sats(:,2) ./ k_brine

qq2(:,:,1) would be the result of co2_sats(:,1) ./ Kb

qq2(:,:,2) would be the result of co2_sats(:,2) ./ Kb

etc.