How do I use the INDEXC Function in SAS?

The Syntax and Arguments of INDEXC

The primary purpose of the INDEXC function is to efficiently identify the location of any individual character defined in a search set within a larger source string. This allows for flexible pattern matching, focusing on character inclusion rather than exact substring matching. When utilized correctly, the function provides a precise numeric index indicating the byte position of the first matching character encountered.

This powerful SAS function utilizes a straightforward syntax that requires two mandatory arguments: the source string to be analyzed and the collection of characters to search for. The function’s structure is defined clearly, ensuring ease of use across various programming contexts within the SAS Data Step.

The basic syntax structure for implementing INDEXC is as follows:

INDEXC(source, excerpt)

Where these arguments fulfill specific roles:

• source: This is the mandatory character string or variable that you intend to analyze. This variable holds the text in which the search operation will take place.
• excerpt: This is the mandatory string containing the specific set of characters the function will search for within the source string. Crucially, INDEXC looks for any single character from this set, not the entire string as a contiguous block.

It is important to remember that SAS character functions are case-sensitive by default, meaning ‘A’ and ‘a’ are treated as distinct characters unless case-insensitive options are employed using related functions like `UPCASE` or `LOWCASE` on the source string prior to execution. For the purpose of the primary INDEXC function, the search is executed based on the exact characters provided in the excerpt argument.

Practical Example: Using the INDEXC Function in SAS

To demonstrate the practical utility of the INDEXC function, consider a scenario where we have a dataset containing customer names, and we need to identify which names contain potentially unusual or specific characters, such as ‘x’, ‘y’, or ‘z’. This type of analysis is common when performing quality checks or preliminary data exploration.

We begin by defining our sample dataset named original_data. This set includes a single character variable, name, which holds various employee names. The code block below illustrates the creation and initial display of this source data:

/*create dataset containing names for analysis*/
data original_data;
    input name $25.;
    datalines;
Andy Lincoln Bernard
Barren Michael Smith
Chad Simpson Arnolds
Derrick Smith Henrys
Eric Millerton Smith
Frank Giovanni Goode
;
run;

/*view dataset contents*/
proc print data=original_data;

Once the source data is confirmed, we can apply the INDEXC function within a new Data Step. Our goal is to search the name variable for any occurrence of the characters ‘x’, ‘y’, or ‘z’. We define the search characters as the string ‘xyz’ in the second argument of the function. The resulting position will be stored in a newly created variable, which we name first_xyz.

The following SAS code executes this search operation and then prints the resulting dataset, allowing us to inspect the generated index values:

/*find position of first occurrence of either x, y or z in name*/
data new_data;
    set original_data;
    first_xyz = indexc(name, 'xyz');
run;

/*view results*/
proc print data=new_data;

Interpreting the Output of the INDEXC Function

The resulting column, first_xyz, provides a precise index (position count) of the first instance where any of the searched characters (‘x’, ‘y’, or ‘z’) appeared in the corresponding name field. Analyzing this output is key to understanding how INDEXC works on a character-by-character basis, contrasting sharply with substring search functions.

A fundamental aspect of interpreting the output is recognizing the significance of the return value zero (0). If INDEXC scans the entire source character string and finds none of the characters specified in the search set, it returns 0. This result is particularly useful for filtering or subsetting data records that do not contain specific invalid or desired characters. Conversely, any positive integer returned signifies the byte position where the match was first identified.

Examining the output table above confirms this behavior. Consider the first observation: “Andy Lincoln Bernard.” The search characters were ‘x’, ‘y’, and ‘z’. The first occurrence of any of these is ‘y’ in “Andy,” which is the fourth character in the string, thus first_xyz correctly returns 4. For the second observation, “Barren Michael Smith,” none of the characters ‘x’, ‘y’, or ‘z’ are present, resulting in a return value of 0. This simple mechanism allows for powerful logical checks against text variables.

The Critical Difference Between INDEX and INDEXC Functions

When working with character manipulation in SAS, developers frequently encounter two similar-sounding functions: INDEXC and INDEX function. While both locate characters within a larger string, their mechanism of operation is fundamentally different. Understanding this distinction is vital to ensure you use the correct tool for your analytical goal.

The standard INDEX function is designed to search for the first occurrence of an entire substring—a contiguous sequence of characters—within the source string. For example, if you search for ‘Smith’, the INDEX function will only return a positive value if the characters S-m-i-t-h appear consecutively in that exact order. It treats the search term as a single, inseparable unit.

In stark contrast, the INDEXC function (where ‘C’ typically stands for Characters or Contains) searches for the first occurrence of any individual character found within the search string argument. If the search string is ‘Smith’, INDEXC scans the source string and stops at the very first occurrence of ‘S’, ‘m’, ‘i’, ‘t’, or ‘h’, regardless of their order or whether they appear together. This distinction dictates which function is appropriate for pattern matching versus exact phrase matching.

Comparative Example: INDEX vs. INDEXC in Practice

To clearly illustrate this functional difference, we will apply both the INDEX and INDEXC functions to our original dataset using the same search term, ‘Smith’. The resulting output variables will highlight how each function interprets the search argument.

We will create two new variables: index_smith using the INDEX function, and indexc_smith using the INDEXC function. Both functions will search the name variable for the string ‘Smith’.

/*create new dataset comparing both functions*/
data new_data;
    set original_data;
    index_smith = index(name, 'Smith');
    indexc_smith = indexc(name, 'Smith');
run;

/*view new dataset*/
proc print data=new_data;

Observing the results in the table provides definitive clarity. The column index_smith utilized the INDEX function and therefore displays the position of the first occurrence of the complete, contiguous substring ‘Smith’ in the name column. Notice that for the first row, “Andy Lincoln Bernard,” the value is 0 because the full word ‘Smith’ is not present.

In contrast, the indexc_smith column, powered by the INDEXC function, displays the position of the first occurrence of any of the individual letters: ‘s’, ‘m’, ‘i’, ‘t’, or ‘h’. In the first row, “Andy Lincoln Bernard,” the letter ‘i’ is the seventh character encountered, causing indexc_smith to return 7. This row perfectly encapsulates the functional difference: one looks for the whole word, the other looks for any component letter.

Advanced Considerations and Related Character Functions

While INDEXC is highly effective for simple character searches, its utility extends into more complex data manipulation tasks. For instance, it is often paired with other SAS functions like `SUBSTR` to extract specific portions of a character string based on the index position returned, or used within conditional statements (`IF-THEN`) to categorize records based on the presence of certain characters.

It is also worthwhile to briefly mention related functions that offer variations on this search logic. The INDEXW function, for example, is designed to search for a word rather than a substring or individual characters, using delimiters to define word boundaries. Furthermore, INDEXANY and INDEXALPHA are specialized functions that leverage similar character set logic to locate specific character classes (like alphabetic characters or numeric digits) without explicitly listing every character in the search argument. Choosing between INDEXC, INDEX, and these specialized variations depends entirely on whether your goal is to find a contiguous block, any character from a set, or a predefined character class.

Summary and Next Steps in SAS Programming

The INDEXC function is an indispensable tool in the SAS programmer’s toolkit, particularly when performing data cleaning, validation, or pattern identification where the presence of individual characters is more important than the location of a complete phrase. By returning the numeric position of the first matched character, or zero if no match is found, it provides a simple yet powerful mechanism for conditional processing and text parsing.

Mastering character functions is a cornerstone of effective SAS programming. The following tutorials explain how to use other common functions in SAS: