Reshape from wide to long

[Meliesa Martin]

I have data in this format

date     comp1_asset    comp1_ profit    comp2_asset    comp2_profit
1/2018    10000        2000        10500        18000
2/2018    12000        2200        11000        19000

The format in which I want is

date       comp    assets    profit
1/2018    comp1    10000    2000
2/2018    comp1    12000    2200
1/2018    comp2    10500    18000
2/2018    comp2    11000    19000

Here is my example data set

clear
input double(ALUMINIUMBAHRAIN ALUMINIUMBAHRAINMARKETVALUE) long ALUMINIUMBAHRAITOTALASSETS double ALUMINIUMBAHRAIBOOKVALUEPE long ALUMINIUMBAHRAICOMMONSHARES double(BAHRAINCARPARKCO BAHRAINCARPARKCOMARKETVA) int BAHRAINCARPARKCOTOTALASSET
.5700000000000001             809.4 1229103 .587 1412731 .16 11.25 12905
.5700000000000001             809.4 1229103 .587 1412731 .16 11.25 12905
              .56 802.3000000000001 1229103 .587 1412731 .16 11.25 12905
              .56 802.3000000000001 1229103 .587 1412731 .16 11.25 12905
              .56 802.3000000000001 1229103 .587 1412731 .16 11.25 12905
              .56 802.3000000000001 1229103 .587 1412731 .16 11.25 12905
.5700000000000001             809.4 1229103 .587 1412731 .16 11.25 12905
.5700000000000001             809.4 1229103 .587 1412731 .16 11.25 12905
.5700000000000001             816.5 1229103 .587 1412731 .16 11.25 12905
              .59             844.9 1229103 .587 1412731 .16 11.25 12905
              .59             844.9 1229103 .587 1412731 .16 11.25 12905
end

[Jorrit Gosens]

This is how far I got.
Note your data, especially the variable names, are quite a mess. I used abbreviated versions of e.g., MARKETVA for market value because some variables contain the full combination MARKETVALUE but others do not.
If other variables in your dataset contain e.g., only the letters MARKETV to indicate market avlue, the code has to be adjusted accordingly.
Note also I assume ALUMINIUMBAHRAI and ALUMINIUMBAHRAIN are the same company, but other than just replacing the name of ALUMINIUMBAHRAI with ALUMINIUMBAHRAIN I have no solution that is more generically applicable.
Also note one variable has no clear indication of what it represents (I named it WHATSTHIS below)
Lastly, note no date was present in you example data so I created a fake one, going on your description.

gen date =_n
ren * v*
ren vdate date
order date, first
reshape long v, i(date) j(somestring) string
gen name=somestring
gen var="bookvalue" if strpos(name, "BOOKVAL")
replace var="marketvalue" if strpos(name, "MARKETVA")
replace var="totalassets" if strpos(name, "TOTALAS")
replace var="commonshares" if strpos(name, "COMMONSH")
replace var="WHATSTHIS" if var==""

replace name = substr(name, 1, strpos(name, "BOOKVAL")-1) if strpos(name, "BOOKVAL")
replace name = substr(name, 1, strpos(name, "MARKETVA")-1) if strpos(name, "MARKETVA")
replace name = substr(name, 1, strpos(name, "TOTALAS")-1) if strpos(name, "TOTALAS")
replace name = substr(name, 1, strpos(name, "COMMONSH")-1) if strpos(name, "COMMONSH")

drop somestring
reshape wide v, i(name date) j(var) string