Programming/Size Related Options
Frequently Used Options
CASE_LENGTH
CASE_LENGTH=#
Shortform: LEN
Specifies the number of columns in the data file for the respondent’s answers. The default is 800 columns, the maximum is 200,000 and the minimum is 10. It can also be specified immediately after the study name without a keyword
EXAMPLE:
[DEMO,15000]
FONE_TEXT_LENGTH
FONE_TEXT_LENGTH=#
Shortform: FONTEX
Specifies the number of columns of text in the phone file. The default is 200, and the maximum is 4900. PREPARE will check that this boundary is not exceeded in PHONE,GET_PHONE_INFORMATION and PHONE,PUT_PHONE_TEXT statements and \[ phone references. If this is set to a number larger than that specified in your phone file, SURVENT will blow.
TEXT_START
TEXT_START=#.#
Shortform: TEX
Defines the text area for each data record which holds all the data from TEXT questions. If the 2nd parameter is not supplied, it defaults to be the rest of the case. If this option is not used, the text area will default to starting on the first column on the next 80 column boundary after the last column accessed anywhere in the questionnaire. There must be at least 100 columns of text area. . The user cannot assign data columns in the text area. The text area must be entirely contained below column 65,535.
WORK_START
WORK_START=#
Shortform: WORK
Allows you to specify the starting location of a data area where the system will not change the next data column pointer after questions where locations are specified. It will also not check for data column overflows or warn you about the next data location being less than this.
You must specify a data column for each question in the work area, or use the keyword “WORK” on a question to put it in the next available work area location.
This is usually used as a place to put data generated from the interview or to store TEXT question pointers or ROTATE seeds if you don’t want them mixed with the rest of the data.
The ending location of the work area depends on the TEXT_START or CASE_LENGTH whichever is encountered first.
Other Options
ANSWER_LENGTH
ANSWER_LENGTH=#
Shortform: ANSLEN
Sets the size of the “answer array” to # bytes. The default is 65,000. The maximum is 1,000,000. This is the approximate number of characters allowable in back-references to questions and is only needed if you have many back-references.
MAXIMUM_GRID_QUESTIONS
MAXIMUM_GRID_QUESTIONS=#
Shortform: MAXGRIDQ
This allows you to have more than the default of 200 questions in a grid. There is no maximum, but if you use this, you may have to increase the amount of core memory SURVENT runs with.
MAXIMUM_JOURNAL_FILE_SIZE
MAXIMUM_JOURNAL_FILE_SIZE=#
Shortform: MAXJSIZE
This keyword can be used in cases where the “response” file for a survey exceeds the default limits. By default, the journal file is 15,000,000 bytes. The absolute maximum you can set it to is 536,870,000 bytes. If you run out of space in the journal file and you receive a warning message, you can update the maximum size by using this keyword.
MAXIMUM_QUESTION_LABELS
MAXIMUM_QUESTION_LABELS=#
Shortform: MAXLAB
Allows you to set the # of labels allowed to a smaller value to impact the size of the compiled questionnaire file (QFF). The default is 64,000 labels. Setting it smaller makes a smaller questionnaire file. 1,000,000 is the maximum allowed. This means that you can do almost use any application without having to split the questionnaire up into multiple CALL questionnaires.
MAXIMUM_QFF_FILE_SIZE
MAXIMUM_QFF_FILE_SIZE=#
Shortform: MAXQFILESIZE
Specifies the maximum size in megabytes for the QFF file. The default is 1 megabyte and the maximum value is 50 megabytes.
MAXIMUM_QUESTION_SIZE
MAXIMUM_QUESTION_SIZE=#
Shortform: MAXQSTSIZE
Allows you to have larger questions, particularly for use with long brand lists, multiple languages or webSurvent. The default size is 64,000 bytes per question. This has the following applications:
10,000 CODE LISTS: This will allow up to 10,000 codes and 750 responses on one question. In particular, this is most useful for a controlling question for a list of cars, etc. where you use FIELD,INCLUDE_EXCLUDE to include or exclude codes from that question.
10,000 LINE DISPLAYS: You can have up to 10,000 lines of display text (take note that browsers may be limited in the amount they can display).
MAXIMUM_PHONE_M
MAXIMUM_PHONE_M=#
Shortform: MAXPHONEM
There is a maximum of 5 market weights that can be changed on a single interview. Updating market weights is resource intensive so you only want to update them when it is necessary. If more than 5 market weights are changed on a single interview SURVENT will get BLOW #. You can use this option to increase the value to more than 5, but do so with caution as it can slow the system down.
MAX_QUOTA_NOW
MAX_QUOTA_NOW=#
Shortform: MAXQUOTANOW
There is a maximum of 5 named quotas that can be updated using the NOW option on a single interview. Since use of that option is resource intensive, you only want to update them when it is necessary. If more than 5 named quotas are updated on a single interview SURVENT will get BLOW #. You can use this option to increase the value to more than 5, but do so with caution as it can slow the system down.
MAXIMUM_QUOTA_NUMBER
MAXIMUM_QUOTA_NUMBER=#
Shortform: MAXQ
Specifies the highest number allowed for numbered quotas. This must be used if numbered quotas are used. # can be a number as small as 100 or as large as 2,156,000,000. The QUO file will grow by 4 bytes for each number specified here.
NOTE: If you specify the maximum number, you will be building a quota file of approximately 40 megabytes in size.
MAX_ODBC_SELECT_SIZE
MAX_ODBC_SELECT_SIZE=#
Specifies the maximum number of bytes that can be in an ODBC select statement. The default is 200,000 bytes.
The issue is that if you had a lot of “text” questions with a lot of text they were overflowing the buffer for the data transport to ODBC. Setting it higher allows more and bigger TEXT questions to be exported.
ROTATE_SEED_WIDTH
ROTATE_SEED_WIDTH=#
Shortform: ROTSEEDWID
Specifies the allowable size of ROTATE,SCRAMBLE statements.
The default is 10, but this lets you set it to the old value of 6 to allow older spec files to retain the old value of six characters. The maximum was increased to 10 to accommodate more questions per rotate. This ensures that all possible combinations of larger rotates will be random. With a six digit random seed, this is not be guaranteed.
SCREEN_LINES
SCREEN_LINES=#
Shortform: SCREEN
Specifies the maximum screen size. The default is the size of the standard screen on the machine you’re compiling on, which is 23 for UNIX and 24 for WINDOWS. If a question has more text lines than #, the system will print whatever it can and prompt to print the rest of the lines on the next screen. This is a terminal only option. Set this to large value like 5000 to avoid any warning messages about the question not fitting on the screen.
SPEC_WIDTH
SPEC_WIDTH=#
Shortform: SPECWID
Specifies the maximum line width of the input specification file and also the length at which the QSP file will wrap lines at. The default and maximum are both 5000. You can set this to smaller values if you want to make sure your QSP file does not get too wide.
TRFILE_DIRECTORY
TRFILE_DIRECTORY=#
Shortform: TRFDIR
This controls the size of the data file case id index when the data file is created by SURVENT. The default is 20,000 entries. The index is used by coding applications to find the record wanted quickly, or in the CLEANIT utility to find cases quickly. If you expect more than 10,000 cases you should increase this value. To remove the case id index for this study, say –TRFILE_DIRECTORY.
NOTE: You can set this for your site in the Survox parmfile.
WEB_PASSWORD_LOCATION
WEB_PASSWORD_LOCATION=#.#
Shortform: WEBPASSLOC
Specifies where in the phone record the password is stored for closed online programs. The default is 51.20.
WEB_USER_NAME_LOCATION
WEB_USER_NAME_LOCATION=#.#
Shortform: TRFDIRLOC
Specifies where in the phone record the user name is stored for closed online programs. The default is 71.20.