How can one do this calculation?

I have a data

Month City Temperature

1 1 25 24 22 19 ...

1 2 21 19 17 16 ...

2 1 23 19 18 17 ..

2 2 20 18 15 13 ..

Then what I like to do is to calculate, for each month, the nth highest temperature from temperature of every city of the month, and then add this info column for each row of this month.

In the above table, for n=2, the second highest temperature in Month 1 is 24, and 20 in Month 2. Thus the output one wants is

Month City 2nd_Hightst_Temp Temperatures

1 1 24 25 24 22 19 ...

1 2 24 21 19 17 16 ...

2 1 20 23 1918 17 ..

2 2 20 20 18 15 13 ..

Thank you in advance.

[Harshit Gupta]

Could you provide a link/attachment of this data?

[Image Analyst]

We'll get to work on it as soon as you attach your data and code to read it in with the paperclip icon. But first read this:

TUTORIAL: How to ask a question (on Answers) and get a fast answer

Answer 1

[dpb]

Assuming you put it into a table as @Steven Lord suggested, then

tHi2=maxk(tData.Temperatures,2,2); tHi2=tHi2(:,2);     % get top two, keep second only
tData=addvars(tData,tHi2,'before','Temperatures','NewVariableNames',{'2nd Highest Temperature'});

ADDENDUM/CORRECTION:

The second optional argument to operate over the second dimension (by rows) was inadvertently omitted...

[alpedhuez]

I was able to solve this issue on my end:

1. for each month, extract temperatures
2. reshape temperatures in a row vector
3. run sort to reorder
4. extract second highest
5. add second highest value to the matrix

[dpb]

See above comment/correction...moral -- read the doc if it isn't quite right!!! I just whiffed on forgetting to write it down when writing online...