This option is available for improving import speed for extremely wide text files that are tab delimited, space delimited, or comma separated. Specify each of the variable names, even if they already exist in the first row of the file.
Variables are either numeric or character. Type a dollar sign ($) following each character variable. Separate names by an empty space or use ranges of variables with a hyphen. For example, specify
to import the first column as character ID, the second column as character Sex, the third column as numeric Age, and the following 100,000 columns as SNP1 to SNP100000 with character type. Note that to classify all the SNP columns in a range as character, you must enclose both the range specification and the $ in parentheses as shown. However, you can often obtain better performance by not using parentheses and instead specifying a common length for the range of character variables in the next parameter (List of Variable Names and Lengths). For example, specify
for the next parameter (List of Variable Names and Lengths) to set a length of 10 for ID, 1 for Sex, 3 for Age, and 2 for all variables SNP1 to SNP500000.
Specifying this parameter and the following parameter (List of Variable Names and Length) for ranges of character variables can significantly reduce import time. The default length of character string is 8. Please define those variables with length larger than 8, or they will be truncated.
Leave this field empty for the system to automatically detect the variable names and types.
|
Type the appropriate line into the List of Variable Names and Types box.
|
Refer to Base SAS documentation on the INPUT statement for additional details.