How to make operation row by row

Hello experts,

this is my code contain:

Alpha_pmp is 400,000 row data

AirTemp is 105200 row data

Tstc is 400,000 row data

below is the formula to calculate FtempN

i can get all the result by using below code:

FtempN =((1 + (Alpha_pmp(1)).* (AirTemp(:,1)-(Tstc(1)))))

if i use above code matlab will return on single row result.

my question is how to to calculate result A by read row by row

for Alpha_pmp and Tstc?

example :

FtempN =((1 + (Alpha_pmp(2)).* (AirTemp(:,1)-(Tstc(2)))))
FtempN =((1 + (Alpha_pmp(3)).* (AirTemp(:,1)-(Tstc(3)))))
FtempN =((1 + (Alpha_pmp(4)).* (AirTemp(:,1)-(Tstc(4)))))
FtempN =((1 + (Alpha_pmp(5)).* (AirTemp(:,1)-(Tstc(5)))))
FtempN =((1 + (Alpha_pmp(6)).* (AirTemp(:,1)-(Tstc(6)))))

until last row.

Thank you.

note : i am newbie.I tried to make some research on matlab web but not luck.

Answer 1

[DGM]

Orient the vectors accordingly and use implicit vector expansion. For example:

% these are all column vectors
% A and C have the same length, but B is shorter
A = (1:20)';
B = (1:10)';
C = A;

% D is now 20x10
% note that B is transposed into a row vector, hence the width of D
D = 1 + A .* (B.' - C)

This implicit array expansion works in R2016b and newer. In older versions, you can use bsxfun():

D = 1 + bsxfun(@times,A,bsxfun(@minus,B.',C))

[Shahid Said]

% these are all column vectors

% A and C have the same length, but B is shorter

A = Alpha_pmp';

B = AirTemp';

C = Tstc';

% D is now 20x10

% note that B is transposed into a row vector, hence the width of D

D = 1 + A .* (B.' - C)

Thanks sir.

but i dun think this is the correct code to get my expected result.

[DGM]

In the example, A, B, C are column vectors because they're analogous to the column vectors you're using.

AirTemp';

isn't a column vector. It's a Mx150k array. I don't know what M is, because you never said how wide AirTemp was.

AirTemp(:,1) 

is a column vector from an array of unknown width. That's all I have to go on.

This should give a 400kx150k array, which will be huge -- problematically huge, but that seems to be what you're trying to calculate.

FtempN = 1 + Alpha_pmp.* (AirTemp(:,1).' - Tstc);

This would give a 150kx400k array instead

FtempN = 1 + Alpha_pmp.' .* (AirTemp(:,1) - Tstc.');

Fwiw, consider that there's 8 bytes per element of a double precision array:

400E3*150E3*8
ans =
   4.8000e+11

[Shahid Said]

this is what i intended to do

I think i dun get the result as expected.

my parameter is

Alpha_pmp have 400k rows

Air Temp have 150k rows

Tstc have 400k rows

so from above I run the code as manual as below :

this is how the code should run in manual

FtempN =1 + ((Alpha_pmp(1)).* (AirTemp(:,1)-(Tstc(1))))

then store in table set1

FtempN =1 + ((Alpha_pmp(2)).* (AirTemp(:,1)-(Tstc(2))))

then store in table set2

FtempN =1 + ((Alpha_pmp(3)).* (AirTemp(:,1)-(Tstc(3))))

then store in table set3

until

FtempN =1 + ((Alpha_pmp(400000)).* (AirTemp(:,1)-(Tstc(400000))))

then store in table set400000

or

create new table nama FtempN_result

FtempN =1 + ((Alpha_pmp(1)).* (AirTemp(:,1)-(Tstc(1))))

then store in FtempN_result column 1

FtempN =1 + ((Alpha_pmp(2)).* (AirTemp(:,1)-(Tstc(2))))

tthen store in FtempN_result column 2

FtempN =1 + ((Alpha_pmp(3)).* (AirTemp(:,1)-(Tstc(3))))

tthen store in FtempN_result column 3

until

FtempN =1 + ((Alpha_pmp(400000)).* (AirTemp(:,1)-(Tstc(400000))))

then store in FtempN_result column 400000

then transpose the table.

so how to run the above code using FOR or any other method ?

my expected result is I will have 400,000 table which each one have 150k rows.

or have 400k of column and 150k of row and transpose the table.

thanks sir for your reply.i am really appreciate.

Answer 2

[KSSV]

Use:

FtempN =(1 + Alpha_pmp.* (AirTemp(1)-Tstc)) ; 

[Shahid Said]

I think i dun get the result as expected.

my parameter is

Alpha_pmp have 400k rows

Air Temp have 150k rows

Tstc have 400k rows

so from above I run the code as manual as below :

this is how the code should run in manual

FtempN =1 + ((Alpha_pmp(1)).* (AirTemp(:,1)-(Tstc(1))))

then store in table set1

FtempN =1 + ((Alpha_pmp(2)).* (AirTemp(:,1)-(Tstc(2))))

then store in table set2

FtempN =1 + ((Alpha_pmp(3)).* (AirTemp(:,1)-(Tstc(3))))

then store in table set3

until

FtempN =1 + ((Alpha_pmp(400000)).* (AirTemp(:,1)-(Tstc(400000))))

then store in table set400000

or

create new table nama FtempN_result

FtempN =1 + ((Alpha_pmp(1)).* (AirTemp(:,1)-(Tstc(1))))

then store in FtempN_result column 1

FtempN =1 + ((Alpha_pmp(2)).* (AirTemp(:,1)-(Tstc(2))))

tthen store in FtempN_result column 2

FtempN =1 + ((Alpha_pmp(3)).* (AirTemp(:,1)-(Tstc(3))))

tthen store in FtempN_result column 3

until

FtempN =1 + ((Alpha_pmp(400000)).* (AirTemp(:,1)-(Tstc(400000))))

then store in FtempN_result column 400000

then transpose the table.

so how to run the above code using FOR or any other method ?

my expected result is I will have 400,000 table which each one have 150k rows.

or have 400k of column and 150k of row and transpose the table.

Thank you sir.