Concatenating 2 vectors into 1 vector with the values adjacent to each other

Hi, I have a three vectors: month, year, and partial pressure of carbon dioxide for 2914 observations. I am trying to use a program to analyze the seasonality of the data, but the program requires that the data be in a 2914x2 (2914 rows, 2 columns) matrix such that column one is the year and month (i.e. February 2012 would be 20121 and June 1985 would be 19856) and column 2 is the partial pressure of carbon dioxide. I've tried countless times to get the data in this format but no matter what I do when I concatenate it forms a 2914x3 matrix. Is there a way to actually concatenate the year and month vectors to get in the format "yearmonth" like 20121 or 19856? Thanks so much!

[Rik]

I would suggest putting data like that in a variable, instead of a row index. You could also consider counting the number of months from your start date instead.

Without knowing what code you used it is difficult to tell you what to change to ensure a two-column array.

[Moey Rojas]

Most of my code has been a derivative of yearmonthCO2=[yr, mon, pCO2]. What I'm trying to do is get a [yearmonth, pCO2] output.

[Stephen23]

The best solution is to use datetime:

https://www.mathworks.com/help/matlab/ref/datetime.html

Answer 1

[Star Strider]

Yes. Use datetime to create them.

Example —

CO2_Mtx = [2012 01 rand; 2012 2 rand; 1985 5 rand; 1985 6 rand];                                % Create Matrix
YM = datetime(num2str(CO2_Mtx(:,[1 2]),'%d %d'), 'InputFormat','yyyy MM', 'Format','yyyyMM');   % Convert ‘Years Month’ To ‘datetimeâ
CO2_Tbl = table(YM,CO2_Mtx(:,3), 'VariableNames',{'YearMth','CO2PP'})                           % Table Of Results

That should work with the data as you have described it. Without your actual matrix however, I cannot be certain.

Producing —

CO2_Tbl =

  4×2 table

    YearMth     CO2PP  
    _______    ________

    201201     0.046171
    201202     0.097132
    198505      0.82346
    198506      0.69483

EDIT —

Added ‘CO2_Tbl’ output.

[Moey Rojas]

Thank you so much, this is such a great explanation. In all of these I was missing the num2str command. Thanks again!

[madhan ravi]

datetime(...) is the way to go!

[Star Strider]

Madhan — Thank you.

I'm deleting my Answer anyway in a few minutes.

[Tommy]

I think this is a great answer and I hope you don't delete it. I'm not quite at the point where I think datetime when I ought to, but I'm getting there, thanks in part to this answer - that's one of the main reasons I spend time on this site! I'd suggest the OP accept this answer.

[Star Strider]

I decided to keep it posted!

The one glitch is that ‘PPCO2’ should have been ‘pCO2’. (And I was a Chemistry major!)

.

Answer 2

[Bjorn Gustavsson]

Why not simply do something like this:

Year_pressure = [year(:) + (month(:)-.5)/12,p_CO2(:)];

Here I've modified your conversion of year+month to "fractional years", mainly because I dont understand how you can map February 2012 to 20121 and June 1985 to 19856, but you can simply replace what I did for the dates with whatever you need for converting the year-month to the format you require.

HTH

[Moey Rojas]

Thank you!

Answer 3

[Tommy]

Should February 2012 map to 20122 rather than 20121? If so, maybe this will do the trick?

month = repmat((1:12)', 5, 1);
year = repelem((1998:2002)', 12, 1);
pCO2 = rand(60, 1);

data = [str2num(sprintf('%d%d\n',[year,month]')), pCO2];

(edit)

>> fprintf('%d %0.3f\n',data')
19981 0.132
19982 0.085
19983 0.232
19984 0.625
19985 0.372
19986 0.130
19987 0.679
19988 0.836
19989 0.662
199810 0.815
199811 0.924
199812 0.314
19991 0.499
...

[Moey Rojas]

Yes, It should have been 20122; that was a bad typo. That worked, thanks!