1.5. SAS PROCEDURES -> Equivalent in R

1.5.2. PROC FREQ - frequency analysis in SAS and R

Produce summaries of grade distribution:

1. Overall frequency
2. By department
3. Cross-tab by department & gender
4. Percentages within group
5. Cumulative insights, pivoting, and formatting

Input Table: students

We’ll assume the following dataset for both SAS and R:

1. Overall Frequency of GRADE

Goal Count how many students got A, B, or C—across all departments.

SAS Code

proc freq data=students;
  tables grade;
run;

• proc freq: Launches the frequency procedure.
• tables grade;: Requests frequency counts for the GRADE column.
• The output includes counts, percentages, cumulative counts and percentages.

R Code

students <- tibble::tibble(
  DEPT = c("Math", "Math", "Math", "Math", "Math", "Physics", "Physics", "Physics", "Physics", "Biology", "Biology", "Biology", "Biology", "Biology", "Biology"),
  GENDER = c("M", "F", "M", "M", "F", "F", "M", "F", "F", "M", "F", "M", "F", "F", "F"),
  GRADE = c("A", "A", "B", "B", "C", "A", "B", "C", "A", "A", "A", "B", "B", "A", "C")
)
students %>% count(GRADE)

• count(GRADE): Counts how many times each GRADE value appears.

Output

2. Grade Frequency by Department

Goal Get grade distribution within each department.

SAS Code

proc sort data=students; by dept;

proc freq data=students noprint;
  by dept;
  tables grade / out=grade_by_dept(drop=percent);
run;

• proc sort: Required before using BY dept; in PROC FREQ.
• noprint: Suppresses printed output and sends results to a dataset.
• out=...: Saves the count output; we drop PERCENT here.

R Code

students %>%
  group_by(DEPT) %>%
  count(GRADE, name = "Count")

• group_by(DEPT): Sets department as the grouping key.
• count(GRADE): Within each department, count how often each grade appears.
• name = "Count": Renames the default n column for readability.

Output

3. Cross-tabulation: GRADE × GENDER within DEPT

Goal See how many males/females got each grade in each department.

SAS Code

proc freq data=students noprint;
  by dept;
  tables grade*gender / out=cross_tab(drop=percent);
run;

• grade*gender: Requests cross-tab between GRADE and GENDER.
• Grouping still applies to each DEPT due to BY dept.

R Code

students %>%
  group_by(DEPT) %>%
  count(GRADE, GENDER, name = "Count")

• count(GRADE, GENDER): Creates 2D frequency tables per department.

Output

4. Add Percentages within Group

Goal Include what percent of each department received each grade.

R Code

students %>%
  group_by(DEPT) %>%
  count(GRADE, name = "Count") %>%
  mutate(Percent = round(Count / sum(Count) * 100, 1))

• mutate(...): Adds a new column Percent.
• Count / sum(Count): Proportion of each grade within its department.
• round(..., 1): Keep results clean to 1 decimal place.

Output

5. Cumulative Metrics (Optional)

students %>%
  group_by(DEPT) %>%
  count(GRADE, name = "Count") %>%
  arrange(DEPT, desc(Count)) %>%
  mutate(
    Percent = round(Count / sum(Count) * 100, 1),
    CumCount = cumsum(Count),
    CumPercent = round(cumsum(Count) / sum(Count) * 100, 1)
  )

• arrange(desc(Count)): Sort grades by most common per department.
• cumsum(Count): Builds cumulative totals.
• CumPercent: Useful for Pareto analysis or targeting thresholds.

6. Wide Format for Reporting

library(tidyr)

students %>%
  count(DEPT, GRADE) %>%
  pivot_wider(names_from = GRADE, values_from = n, values_fill = 0)

• pivot_wider(): Converts long → wide format for tabular reports.
• values_fill = 0: Ensures missing values show as 0.

Output

7. Visualization: Stacked Bar Chart

library(ggplot2)

students %>%
  count(DEPT, GRADE) %>%
  ggplot(aes(x = DEPT, y = n, fill = GRADE)) +
  geom_col(position = "stack") +
  labs(title = "Grade Distribution by Department", y = "Student Count")

• geom_col() uses pre-counted data.
• fill = GRADE stacks by grade per department.
• Helpful for presentations and spotting trends visually.

Output:

**Summary Table: SAS vs R**

**Resource download links**

1.5.2.-PROC-FREQ-frequency-analysis-in-SAS-and-R.zip