ETBCONV Version 1.01
8 March 1990
A program to convert CASP (CDC Anthropometric Software Package) ETB
files to dBase III, ASCII, comma-delimited,
or Epi Info files.
Supported and distributed by:
Nutrition Division
Centers for Disease Control
1600 Clifton Road NE, MS A08
Atlanta, GA 30333
USA
Phone Number: (404)639-3671
Fax Number: (404)639-3037
and
Nutrition Unit
World Health Organization
1211 Geneva 27
Switzerland
Written in Turbo Pascal by David A. Foster, Department of
Epidemiology, University of Michigan.
ETBCONV is distributed with Epi Info Version 5 and CASP as a
utility program to convert CASP files into other file types. This
program can run in the menu mode or command-line mode. The menu mode
will be described first followed by a description of the command-line
mode.
Menu Mode
To use the menu mode, enter "ETBCONV" (without the quotation
marks) at the DOS prompt. A menu will appear on the screen
requesting information for:
(1) .ETB file name
(2) Output file name
(3) Output file type - either D (for dBase), A (ASCII), C
(comma-delimited), or E (Epi Info)
(4) Whether or not to keep deleted records
Provide the appropriate responses for the questions.
The extensions for the different output file types are:
(1) .DBF for dBase files
(2) .DAT for ASCII files - in addition, the field names and
column positions are in a separate file with the extension .AHD
(3) .PRN for comma-delimited files - in addition, the field
names are in a separate file with the extension .CHD
(4) .REC for Epi Info files
The field names for dBase, ASCII, and comma-delimited files
(excluding the added fields) and their definitions are:
ID Identification information
SEX Sex
AGE Age in months
HT Height in centimeters
WT Weight in kilograms
HAP Height-for-Age Percentile
HAZ Height-for-Age Z-score
HAM Height-for-Age Percent-of-Median
WAP Weight-for-Age Percentile
WAZ Weight-for-Age Z-score
WAM Weight-for-Age Percent-of-Median
WHP Weight-for-Height Percentile
WHZ Weight-for-Height Z-score
WHM Weight-for-Height Percent-of-Median
DOB Date of Birth
DOV Date of Visit/weight and height measurements
FLAG Record flag (more information on this below)
For added fields, the field names are changed if they contain
invalid characters for dBase or Epi Info file field names. If the
first character is an invalid character, it will be changed to the
letter "Z". For example, if a field name starts with the character
"@", the "@" will be changed to a "Z". Other invalid characters are
changed to an underscore (i.e., "_"). Blanks within an .ETB added
field name are removed.
Date fields in ASCII and comma-delimited files will have the
format YYYYMMDD, e.g., 19880130 would be the date for January 30,
1988.
When the option to copy deleted files is set to "Yes", records
transferred to dBase and Epi Info files will be marked for deletion
within those databases.
A "Flag" field is added to each record. This added field is used
to identify records that, in the majority of instances, have missing
information or data errors (such as weight entered as 55 kgs rather
than 5.5 kgs). It would be extremely rare to find individuals who
would have such extreme anthropometric values.
The criteria for "flagging" records is based on Z-score values are
as follows:
Index Minimum Maximum
Height-for-Age -6.00 6.00
Weight-for-Height -4.00 6.00
Weight-for-Age -6.00 6.00
Two additional criteria for "flagging" records are combinations of
data items:
Both HAZ>3.09 and WHZ<-3.09
or
Both HAZ<-3.09 and WHZ>3.09
We recommend that all "flagged" records be verified for accuracy.
A coding scheme for identifying which index or indices are flagged is
shown below:
Flag Index Flagged
Code HAZ WHZ WAZ Notes
-------- ----- ------ ----- ----- -----------
0 No indices flagged
1 Y HAZ flagged only
2 Y WHZ flagged only
3 Y Y Both HAZ and WHZ flagged
4 Y WAZ flagged only
5 Y Y Both HAZ and WAZ flagged
6 Y Y Both WHZ and WAZ flagged
7 Y Y Y All three indices flagged
-------- ----- ------ ----- ----- -----------
Y=Indice flagged
In analyzing anthropometry data, it is recommended that certain
indices be set to missing values (and therefore excluded from
analyses) based on the coding in the Flag field. For example,
analyses of Weight-for-Height values (whether in terms of Z-scores,
percentiles, or percent-of-median) should exclude records that have
the codes 2, 3, 6, or 7.
A special note on Epi Info files
All of the comments that follow refer to Epi Info Version 5. The
Epi Info file created by ETBCONV can be analyzed within the Epi Info
ANALYSIS module. If you want to edit or add records to the file,
please read the chapter related to Nutritional Anthropometry in the
Epi Info manual. Also note that you must have a Check (.CHK) file in
order to perform the anthropometric calculations within Epi Info.
The .CHK file distributed with Epi Info (MEASURE.CHK) can be used.
First, copy the MEASURE.CHK file to one with your Epi Info file name.
For example, if the Epi Info file name is RURAL.REC, enter the
following command in the Epi Info subdirectory:
COPY MEASURE.CHK RURAL.CHK
This will copy the MEASURE.CHK file information into a new file
called RURAL.CHK. You also need to copy MEASURE.BAT to a file called
RURAL.BAT. After you have created the RURAL.BAT file, edit the file
in the Epi Info EPED module. The contents of the RURAL.BAT file
should be as follows:
MARK
DOMEAS
ENTER MEASURE
REMOVE
Change the third line from ENTER MEASURE to ENTER RURAL (which can
be done in Epi Info's ENTER module). Save the file and return to the
DOS prompt and enter:
RURAL
This should load in the RURAL file and allow for editing, etc.
However, you will probably want to modify the .CHK file to change the
skip patterns if there are added fields. Please consult the Epi Info
documentation on how to modify .CHK files. There is also a sample
data set supplied with Epi Info called NUTRI that may serve as an
useful example for modifying your .CHK file.
The Epi Info file created by ETBCONV has the following format:
Name Identification information
SEX Sex
AGE Age in months
BIRTHDATE Date of Birth (mm/dd/yy)
YRS Age in years
VISITDATE Date of Visit/weight and height measurements (mm/dd/yy)
WEIGHT Weight in kilograms
LBS Weight in pounds
OZS Weight in ounces
HAP Height-for-Age Percentile
HAZ Height-for-Age Z-score
HAM Height-for-Age Percent-of-Median
HEIGHT Height in centimeters
WAP Weight-for-Age Percentile
WAZ Weight-for-Age Z-score
WAM Weight-for-Age Percent-of-Median
FT Height in feet
IN Height in inches
WHP Weight-for-Height Percentile
WHZ Weight-for-Height Z-score
WHM Weight-for-Height Percent-of-Median
FLAG Record flag (more information on this below)
You may want to modify the file to meet your needs.
The menu mode entry provides full keyboard editing capabilities
such as insert/overwrite, destructive backspace, character delete at
cursor position, and cursor positioning with arrow keys.
Command-Line Mode
The command-line mode pf ETBCONV requires the entry of 4
parameters at the DOS prompt:
(1) the name (without extension) of the ETB file
(2) the name (without extension) of the output file
(3) the type of output file:
D = dBase III Plus
A = ASCII file (fixed column)
C = comma-delimited ASCII file
E = Epi Info file
(4) the option to pass or retain deleted records
Y = yes, pass deleted records
N = no, retain deleted records
Example 1 of the command-line mode:
etbconv infile outfile d y
The example 1 command would use infile.etb to produce
outfile.dbf (a dBase file). Deleted records would be passed.
Example 2 of the command-line mode:
etbconv infile outfile a n
The example 2 command would use infile.etb to produce
outfile.dat (an ASCII file) and the corresponding file structure
information in the file outfile.ahd. If the third parameter in
the command line in Example 2 were a "c" (instead of "a"), an
ASCII comma-delimited file would be produced with the name
outfile.prn, and the corresponding header file would be
outfile.chd.
Comparison of Menu and Command-Line Modes
The menu mode provides a higher level of protection than command-
line mode in that the menu mode provides:
(1) an overwrite warning on existing output files
(2) an interactive problem explanation for data entry steps
(3) record processing number report during file generation
(4) output report
Both command-line and menu modes report inability to find the ETB
file. The menu mode provides an Abort/Continue option at that point,
command-line mode reports the error and terminates. The header files
created when ASCII or comma-delimited output types are specified
contain the field names and positions. To read the header file, you
can use the DOS "TYPE" command and to print the header file use the
DOS "PRINT" command (see your DOS manual for further information on
these DOS commands).
Problems or Suggestions
Should you have any suggestions or problems concerning the
software, please contact the Nutrition Division, Centers for Disease
Control (address, phone number, and FAX number listed above). For
problems, please send the following information:
(1) A description of the problem
(2) A copy of the screen where the problem occurred (using Prt
Scrn)
(3) The type of computer, monitor, and graphics card (if any)
(4) A copy of the autoexec.bat and config.sys files (if any)
Distribution Conditions
NON-WARRANTY. ETBCONV is provided "as is" and without any
warranty expressed or implied. The user assumes all risks of the use
of ETBCONV. ETBCONV may not run on your particular hardware/software
configuration. We bear no responsibility for any mishap or economic
loss resulting therefrom the use of this software.
COPYRIGHT CONDITIONS. You may make and distribute copies of
ETBCONV provided that there is no material gain involved.
USE AT YOUR OWN RISK. All risk of loss of any kind due to the use
of ETBCONV is with you, the user. You are responsible for all
mishaps, even if the program proves to be defective.
etbconv.doc
|
