File Utilities: Converting, Reformatting and More

Converting File Formats
- StatTransfer
  - Software that can convert among over 20 data formats
  - Available in CUIT lab PC's and DSSC workstations
- Options within SAS, Stata and SPSS
  
  Although StatTransfer's versatility make it the preferred way to convert
  files, when it is not available there are other options.
  - SPSS can read
    SAS, Stata, Excel, dBase, and ASCII fixed format or delimited files
  - Stata can read
    Excel, SAS export files and ASCII fixed format or delimited files
  - SAS can read
    Excel, dBase and ASCII fixed format or delimited files
File Transfer Protocol

FTP is a utility for transferring files between a local host (your computer)
and remote hosts (some other computer).
- FTP on Windows (WINSCP),
  on CUIT and DSSC PCs in the Network Apps folder of Program Menu
- FTP on Macintosh (FUGU)
- FTP on UNIX
Frequently Used Data Reformatting Tools

As per the EDS consulting guidelines, staff can help you get your data ready for use.
This includes procedures like with merging, subsetting, joining, recoding.
Here are examples of techniques we use.
- Excel: converting formulas into values
  - Highlight the cells you want to copy.
  - Copy the cells (ctrl-C or "copy" from the Edit menu).
  - On the Edit menu, choose Paste Special.
  - Click on "Values" under Paste, and then click OK.
- Stata: taking samples, use the "sample" command.
  - a 10% sample -- sample 10
  - sample of a subset -- by sex: sample 10
  - a number sample -- sample 100, count (for sample with 100 records)
  - a contest winner -- sample 1, count
  - sample based on record number -- keep if mod(_n,10)==0
    ( every 10th record)
- Stata: Strings in to Numbers use the "destring" command
  - In codebook check that valid values do not include alpha characters
  - to destring variable age_wed -- destring age_wed
  - The force options generates missing for alpha characters
- Stata: add leading zeros to a number.
  - Convert number variable, say ID_N to a string variable called ID_S
  - gen str3 ID_Z = string(ID,"%03.0f")
  - values like, 1 11 103 will become 001 011 103