Destring error: r(134) too many values

[Martin Imelda Borg]

Hi I am trying to destring my values which contain a mixture of numbers and letters - see below

ID
0000366CAX878
00000HG78888
8787781HG0101

As it's a mixture of numbers and letters I am using the 'destring' function

Code:
destring ID, gen(ID_N) force

However, this generated a couple of thousand missing values - I have 1mill records

The same amount of missing values were generated with this code:
gen (ID_N) = real(ID)

How do I go about this problem?
____

Of course, I didn't use the encode option as I have 1mil records and although this would be the wrong function as it the table ID includes letters there are 'too many values'

[Rich Goldstein]

-destring- should only be used when the variable is (basically) numeric; see

h encode

if you really need a numeric version; however, for virtually all purposes within Stata, the ID can be a string variable so why do you want it to be numeric?

[Carlo Lazzaro]

Martin:
you may want to try:

. egen wanted=group( ID )

. list

     +------------------------+
     |            ID   wanted |
     |------------------------|
  1. | 0000366CAX878        2 |
  2. |  00000HG78888        1 |
  3. | 8787781HG0101        3 |
     +------------------------+

[Nick Cox]

If you need a numeric identifier, then #3 is the only strategy if destring and encode will not deliver the goods. It is equivalent to

sort ID
gen wanted = sum(id != ID[_n-1])