Free A00-231 practice test: 30 questions with answers

NODUPKEY checks only the BY variables, so one row per customer ID is kept even when other fields differ. NODUP compares every variable so rows with different addresses or phones would both remain, DESCENDING changes the sort direction, and OUT= only names the output data set.

The sum statement automatically retains its variable across iterations, initializes it to zero, and ignores missing values in the expression. It does not need a BY variable or a RETAIN statement, and it is a DATA step statement.

BY-group processing with FIRST. and LAST. variables lets the step restart the total for each department and write one row when the last employee in that department is reached. A WHERE filter keeps rows rather than summing them, PROC TRANSPOSE reshapes rows without adding them, and LENGTH only controls storage width.

SAS sets the length of a new character variable from its first appearance in the DATA step, so the two-character value 'No' fixes the length at two. The longer value is truncated to fit, which is why a LENGTH statement placed before the first reference is recommended. The other options describe automatic extension, a 200-character default, or an error that SAS does not raise here.

DO UNTIL evaluates its condition at the end of each pass, so the body always runs at least once. DO WHILE tests first and can run zero times, an iterative DO runs according to its counter and can also run zero times, and an IF-THEN statement is not a loop.

BY store defines one output row per store, ID quarter names the new columns from the quarter values, and VAR sales supplies the values placed in those columns. The other choices assign the variables the wrong roles, such as using store as the ID or making the grouping variable the source of the values.

Macro variable references resolve inside double quotes but not inside single quotes, so only the double-quoted &region version shows West. Leaving out the ampersand prints the word region, and &regions refers to a different, undefined macro variable.

A %LET outside a macro creates a global macro variable that remains available for the rest of the session. A %LET inside a macro definition creates a local variable by default, and %LET never creates a column in a data set.

The WORK library is the default temporary library, and its data sets are deleted when the SAS session ends. A permanent library keeps its data sets after the session, an exported CSV file is a copy outside SAS, and SASHELP holds SAS-supplied sample data rather than session output.

In a two-level name, the first part is the library reference that points to a library and the second part is the member name. Reversing the two levels, limiting names to views, or restricting them to WORK and SASHELP are all incorrect.

A MERGE with a BY statement performs a match-merge on the sorted key, bringing customer values onto each matching order row. SET stacks the tables vertically instead of adding columns, a MERGE without BY pairs rows by position rather than by key, and PROC APPEND adds rows to a base table.

SAS stores a date as the number of days since January 1, 1960, so the value 0 is that date. January 1, 1970 is the Unix epoch, December 31, 1959 is the day before the SAS base date (value -1), and January 1, 1900 is not the SAS base date.

The OBS= data set option stops reading after the given observation number, so only the first 100 observations are read. FIRSTOBS= begins reading at a later observation, KEEP= selects variables rather than rows, and a WHERE= test on id values filters by content rather than by position.

PROC APPEND adds the observations of the DATA= data set to the end of the BASE= data set, so the large history is not copied. A SET step rewrites every observation into a new data set, while PROC SORT with NODUPKEY and a BY-based MERGE do not append rows.

PROC CONTENTS reports variable names, types, lengths and observation counts for a data set, and PROC DATASETS lists the members of a library and can show their contents. PROC SORT reorders observations and PROC MEANS calculates statistics, so neither is used to investigate a library's structure.

The IN= data set option creates a temporary variable that is 1 when that input supplied the current observation and 0 otherwise, so it can be tested in an IF statement. KEEP= selects variables, a BY statement on sales reorders the rows instead of tracking their origin, and PROC APPEND does not record where each row came from.

Each separate IF is tested independently, so a score of 95 satisfies all three conditions and the last assignment, C, overwrites the A. An IF-THEN with ELSE IF stops at the first true condition and gives 95 the grade A, while changing the operators, moving code to another step, or changing the data type does not stop the overwriting.

Syntax errors break the language rules, so SAS reports them in the log and the affected step does not run as intended. Logic errors are valid code that does not match the programmer's intent, so the step may run normally and produce wrong values.

SAS ends a statement only at a semicolon, so the two assignments merge into one invalid statement and the log shows an error. SAS does not insert semicolons, silently skip statements, or blank out every variable in the step.

In comparisons SAS treats a missing numeric value as smaller than any number, so missing income < 50000 is true and band is set to Low. The other options assume missing values make the comparison false, cause a syntax error, or remove the observation, which SAS does not do here.

PUT statements show variable values as the step executes, and a small OBS= sample makes it easier to trace each observation through the logic. Removing RUN merges the next statements into the current step rather than fixing the logic, and NOTE messages about missing values often reveal the cause of wrong results.

This note means that an arithmetic or function operation received a missing operand and returned a missing result, and the step still completes. The other options describe a failed step, an undefined macro reference, or a sort-order problem, none of which this note reports.

When a numeric field holds text that cannot be read as a number, SAS sets that variable to missing and sets the automatic variable _ERROR_ to 1 for that iteration. Text cannot be stored in a numeric variable, the value is not converted to zero, and the step does not stop at the first invalid row.

NOOBS in PROC PRINT suppresses the observation-number column. NOCUM and NOPERCENT are PROC FREQ options for frequency tables, and NODUPKEY is a PROC SORT option, so none of them affects the Obs column in PROC PRINT.

The TABLES statement with an asterisk requests a two-way crosstabulation in a single table. BY produces separate listings for each group, CLASS is used by summary procedures such as PROC MEANS, and WEIGHT changes how observations are counted rather than building a cross-tab.

A user-defined format changes only how values are displayed, so the stored M and F remain unchanged. RENAME= changes variable names rather than values, LENGTH controls storage width, and KEEP= selects variables, so none of them displays new labels for the codes.

A TITLE statement stays in effect across later steps until another TITLE is submitted or the titles are cancelled. Titles are not stored as data set variables, and the end of a PROC step does not remove them.

An ODS destination must be closed with ODS HTML CLOSE to finish the output file. ENDSAS ends the whole session, RUN and QUIT only end individual steps while the destination stays open, and OPTIONS NOSOURCE changes only the log display.

PROC MEANS with a CLASS statement computes the requested statistic for each combination of CLASS values, giving one line per region. PROC PRINT lists every observation, PROC FREQ counts combinations rather than averaging sales, and NODUPKEY keeps one row per region without computing an average.

ODS TRACE ON writes the path and name of each output object to the log, and those names can then be used in an ODS SELECT statement. ODS EXCLUDE ALL suppresses every table, ODS HTML CLOSE ends a destination, and NOSOURCE affects only how program lines appear in the log.