Dear List Members,
I downloaded both the Windows and (Ubuntu) Linux new core programs to test. So far, I got to look at the Windows version.
The test data used was some 5000 records created for a Juvenile Criminal Justice Research Project. The data was generated by a handy program of Christian Maas http://www.chmaas.handshake.de/delphi/freeware/freeware.htm -- Random file generator, command line version. The command In a BAT file was:
RndFileC.exe /R=5000 /Z=YYYY/MM/DD /N=8 /F=1,; /N=9 /F=1,; /C=13 /F=1,; /C=1 /F=1,; /C=15 /F=1,; /C=25 /F=1,; /C=13 /F=1,; /C=2 /F=1,; /C=9 /F=1,; /C=2 /F=1,; /Y=1945,1960 /F=1,; /C=1 /F=1,; /Y=1950,1965 /F=1,; /C=3 /F=1,; /N=5 jpomas.dat
and a few lines of the data in `jpomas.dat' looks like:
00002209;710069725;KEYLKIESBSHSL;W;CESFMALUBIGMGQE;DYMWHGTMIIHDLBYYWBUNMPOZZ;AHYDWEHMBMGVO;DR;RKKBMUEVH;JS;1959/02/08;E;1965/05/12;QAT;46251 00013938;240660830;TJXYNVNHMDRKP;K;ICTRFDWLWCMDGXN;SKXNRZEWIRNVFJXGAFZEVJTLJ;MRUPUDXLXGTDL;KX;DKEMPMZMV;EN;1952/03/02;X;1951/11/24;TIN;48246 00047184;652253478;IDVWGDMZKVLTG;E;AJUMDXOHBJYUCAN;XIRQVGNPYOHQVCKNCUBLVZWRR;ZQKVPGACWHGRV;AB;WIYVZUPJB;XB;1960/05/16;P;1952/08/18;WGO;36580 00048212;488662713;RBOQXPCOSZRMC;S;GPULYLUGZSPUGOW;IDCXHZDBLYJKLGSIIMUHQZYGQ;LXYLTQEPEWDKR;RF;GAEFQRKBY;BE;1959/06/04;J;1957/09/30;VIH;59715 00049725;109094902;IHUIAKMQENBPE;C;OLULCBDSLRNPBEV;AWEFXLALVDECXRRPNZVUPBBFB;YVPXBBIXFYKQX;ST;ALULZKOKY;CO;1957/07/09;Q;1959/10/04;YYF;52178 00063364;024917810;LROXMBYVMEJMM;T;HUQSESVIYXOMPAE;ADFSXIPCDIVVMSTZEATETKJAC;BVILTKROGTHKB;HL;MBGWVLVUF;XZ;1949/07/29;N;1958/04/19;CHE;75683
Data wa imported into the stable EpiData 3.1 (270108) with dictionary:
DATA FILE: C:\Documents and Settings\Owner.YOUR-E92F6775CF\My Documents\jpo_lc\rand-jpo.rec File label: [none]
File size: 645 kb Last revision: 1. May 2008 1:04 PM Number of fields: 15 Number of records: 5000 Checks applied: Yes (Last revision 1. May 2008 1:05 PM)
Fields in data file:
No. Name Variable label Field type Width Checks -------------------------------------------------------------------------------- 1 i1 Client Ident Text 8 Key unique 1 Code Must enter
2 i2 Soc Security Nbr Text 9
3 i3 First Name Text 13
4 i4 Middle Initial Upper-case text 1
5 i5 Last Name Text 15
6 i6 Local Address Text 25
7 i7 City Text 13
8 i8 State Upper-case text 2
9 i9 Zip Code Upper-case text 9
10 i10 County Code Upper-case text 2
11 d1 Date of Birth Date (mdy) 10
12 v1 Gender Upper-case text 1
13 i11 Date Entered Today (mdy) 10
14 i12 Entered by Upper-case text 3
15 mid Auto ID ID number 5
--------------------------------------------------------------------------------
Since I don't have either Stata or a program which uses dBase, the best I could do was export the EpiData file to Stata and then read it back in. Seems `i4' and `v1' were blank and the dates were an assortment of +/- numbers (probably how dates are kept in Stata). The check file (as originally read from the EpiData 3.1 database was all right, but this was a very simple one and not UTF-8.
Other thoughts: maybe a SQLite import/export http://www.sqlite.org/ later on.
Pete Geddes
epidata-list@lists.umanitoba.ca wrote:
Dear All <...> It would be constructive for the development of the Core module if users could find some time to challenge the core-module-test application in Linux or Windows. See www.epidata.dk/testing.php Report success and/or failures to the list.
regards Jens Lauritsen EpiData Association <...>