Separate data

[saptorshee chakraborty]

Dear Stata Users,


I have a dataset of WINDSPEED (value in the dataset below) for each Lat/Long combination for every month of 2023 starting from Jan till Dec (i.e.,for January the data starts with 90 Lat 0 Long combination and finishes at -90 Lat 359.75 Long combination then the data for February starts and so on.)

Is it possible to separate the Value variable according to Months for each Lat/Long combination.

* Example generated by -dataex-. For more info, type help dataex
clear
input float(latitude longitude value)
90     0 7.616347
90   .25 7.616347
90    .5 7.616347
90   .75 7.616347
90     1 7.616347
90  1.25 7.616347
90   1.5 7.616347
90  1.75 7.616347
90     2 7.616347
90  2.25 7.616347
90   2.5 7.616347
90  2.75 7.616347
90     3 7.616347
90  3.25 7.616347
90   3.5 7.616347
90  3.75 7.616347
90     4 7.616347
90  4.25 7.616347
90   4.5 7.616347
90  4.75 7.616347
90     5 7.616347
90  5.25 7.616347
90   5.5 7.616347
90  5.75 7.616347
90     6 7.616347
90  6.25 7.616347
90   6.5 7.616347
90  6.75 7.616347
90     7 7.616347
90  7.25 7.616347
90   7.5 7.616347
90  7.75 7.616347
90     8 7.616347
90  8.25 7.616347
90   8.5 7.616347
90  8.75 7.616347
90     9 7.616347
90  9.25 7.616347
90   9.5 7.616347
90  9.75 7.616347
90    10 7.616347
90 10.25 7.616347
90  10.5 7.616347
90 10.75 7.616347
90    11 7.616347
90 11.25 7.616347
90  11.5 7.616347
90 11.75 7.616347
90    12 7.616347
90 12.25 7.616347
90  12.5 7.616347
90 12.75 7.616347
90    13 7.616347
90 13.25 7.616347
90  13.5 7.616347
90 13.75 7.616347
90    14 7.616347
90 14.25 7.616347
90  14.5 7.616347
90 14.75 7.616347
90    15 7.616347
90 15.25 7.616347
90  15.5 7.616347
90 15.75 7.616347
90    16 7.616347
90 16.25 7.616347
90  16.5 7.616347
90 16.75 7.616347
90    17 7.616347
90 17.25 7.616347
90  17.5 7.616347
90 17.75 7.616347
90    18 7.616347
90 18.25 7.616347
90  18.5 7.616347
90 18.75 7.616347
90    19 7.616347
90 19.25 7.616347
90  19.5 7.616347
90 19.75 7.616347
90    20 7.616347
90 20.25 7.616347
90  20.5 7.616347
90 20.75 7.616347
90    21 7.616347
90 21.25 7.616347
90  21.5 7.616347
90 21.75 7.616347
90    22 7.616347
90 22.25 7.616347
90  22.5 7.616347
90 22.75 7.616347
90    23 7.616347
90 23.25 7.616347
90  23.5 7.616347
90 23.75 7.616347
90    24 7.616347
90 24.25 7.616347
90  24.5 7.616347
90 24.75 7.616347
end

[Carlo Lazzaro]

Saptorshee:
I would first crate a variable that group lat and long:

 egen wanted=group( latitude longitude)

Then you can -sort- according to months.

[Nick Cox]

In what you show there is one and only one windspeed value for each (latitude, longitude) pair, so you would end up with as many variables as observations. It's not clear what that would achieve.

On a different level: latitude 90 N is the North Pole and latitude 90 S is the South Pole and at either pole longitude is indeterminate.

So, I suppose it makes sense that windspeed is reported the same for different longitudes.

[saptorshee chakraborty]

Carlo Lazzaro Thank you very much, but could you let me know after sorting the new variable ' wanted'. How can I separate the data and create new variables ' value_jan, value_feb,...., value_dec' for each lat/long combination.


Regards

[Hemanshu Kumar]

You might start with something like

gen byte new_month = (latitude == 90 & longitude == 0)
gen cumul_month = sum(new_month)

and go from there. The details would depend on what other variables you have to denote year etc.

[Carlo Lazzaro]

Saptorshee:
things would be easier is you share an example that includes your -month- variable. Thanks.

[saptorshee chakraborty]

Nick Cox Thank you for your reply, my post might not have been clear. The data is for each Lat/Long combination monthly average. Since the data was from GRIB file (fig contains data for year 2023) converted to CSV, data is sorted for month one after another in the column value. I want to seperate the value according to month for each LAT/LONG combination.

Regards

[saptorshee chakraborty]

Hemanshu Kumar Thank you very much. This works perfectly.


Regards

[saptorshee chakraborty]

Hemanshu Kumar by any chance is there a way to rearrange the COLUMN "value" according to "latitude", "longitude" and "cumul_month_1", "cumul_month_2","cumul_month_3",....,"cumul_month_ 12"


Regards

[Hemanshu Kumar]

Could you explain what you mean, perhaps showing a small example of what your final dataset should look like?

[saptorshee chakraborty]

Sorry for the confusion.... before you had suggested the way to use the to seperate the data according to months. Is it somehow possible to break the "value" column according to each LAT/LONG combination and "cumul_month"

* Example generated by -dataex-. For more info, type help dataex
clear
input float(latitude longitude value) byte new_month float cumul_month
90     0 7.616347 1 1
90   .25 7.616347 0 1
90    .5 7.616347 0 1
90   .75 7.616347 0 1
90     1 7.616347 0 1
90  1.25 7.616347 0 1
90   1.5 7.616347 0 1
90  1.75 7.616347 0 1
90     2 7.616347 0 1
90  2.25 7.616347 0 1
90   2.5 7.616347 0 1
90  2.75 7.616347 0 1
90     3 7.616347 0 1
90  3.25 7.616347 0 1
90   3.5 7.616347 0 1
90  3.75 7.616347 0 1
90     4 7.616347 0 1
90  4.25 7.616347 0 1
90   4.5 7.616347 0 1
90  4.75 7.616347 0 1
90     5 7.616347 0 1
90  5.25 7.616347 0 1
90   5.5 7.616347 0 1
90  5.75 7.616347 0 1
90     6 7.616347 0 1
90  6.25 7.616347 0 1
90   6.5 7.616347 0 1
90  6.75 7.616347 0 1
90     7 7.616347 0 1
90  7.25 7.616347 0 1
90   7.5 7.616347 0 1
90  7.75 7.616347 0 1
90     8 7.616347 0 1
90  8.25 7.616347 0 1
90   8.5 7.616347 0 1
90  8.75 7.616347 0 1
90     9 7.616347 0 1
90  9.25 7.616347 0 1
90   9.5 7.616347 0 1
90  9.75 7.616347 0 1
90    10 7.616347 0 1
90 10.25 7.616347 0 1
90  10.5 7.616347 0 1
90 10.75 7.616347 0 1
90    11 7.616347 0 1
90 11.25 7.616347 0 1
90  11.5 7.616347 0 1
90 11.75 7.616347 0 1
90    12 7.616347 0 1
90 12.25 7.616347 0 1
90  12.5 7.616347 0 1
90 12.75 7.616347 0 1
90    13 7.616347 0 1
90 13.25 7.616347 0 1
90  13.5 7.616347 0 1
90 13.75 7.616347 0 1
90    14 7.616347 0 1
90 14.25 7.616347 0 1
90  14.5 7.616347 0 1
90 14.75 7.616347 0 1
90    15 7.616347 0 1
90 15.25 7.616347 0 1
90  15.5 7.616347 0 1
90 15.75 7.616347 0 1
90    16 7.616347 0 1
90 16.25 7.616347 0 1
90  16.5 7.616347 0 1
90 16.75 7.616347 0 1
90    17 7.616347 0 1
90 17.25 7.616347 0 1
90  17.5 7.616347 0 1
90 17.75 7.616347 0 1
90    18 7.616347 0 1
90 18.25 7.616347 0 1
90  18.5 7.616347 0 1
90 18.75 7.616347 0 1
90    19 7.616347 0 1
90 19.25 7.616347 0 1
90  19.5 7.616347 0 1
90 19.75 7.616347 0 1
90    20 7.616347 0 1
90 20.25 7.616347 0 1
90  20.5 7.616347 0 1
90 20.75 7.616347 0 1
90    21 7.616347 0 1
90 21.25 7.616347 0 1
90  21.5 7.616347 0 1
90 21.75 7.616347 0 1
90    22 7.616347 0 1
90 22.25 7.616347 0 1
90  22.5 7.616347 0 1
90 22.75 7.616347 0 1
90    23 7.616347 0 1
90 23.25 7.616347 0 1
90  23.5 7.616347 0 1
90 23.75 7.616347 0 1
90    24 7.616347 0 1
90 24.25 7.616347 0 1
90  24.5 7.616347 0 1
90 24.75 7.616347 0 1
end

[Nick Cox]

help separate

[Hemanshu Kumar]

What you are requesting in #11, if I am understanding it correctly, would create millions of new variables, and would exceed Stata's limits on the number of variables. I am also quite certain it is entirely unnecessary. Could you tell us the broader problem you are trying to solve, so that better solutions can be found?

Note:

1. See XY problem
2. It is best not to address your question only to me or any specific individual. Others may well be able to answer your questions better and sooner.