Creating Sample Data Made Easy: SAS vs R – A Friendly Comparison

Creating sample data is essential for testing, simulation, and analysis. Both SAS and R provide versatile methods to generate sample datasets. In this post, we'll compare how to create sample data side by side in SAS and R, offering more than five methods with examples and explanations.

Method 1: Using Basic Data Structures

SAS

In SAS, the DATA step is the most basic way to create a sample dataset.

/* SAS Example: Basic Data Step */
DATA sample_data;
    input ID Name $ Age Height Weight;
    DATALINES;
    1 John 25 68 150
    2 Jane 30 65 120
    3 Jim 22 70 180
    4 Jill 28 64 135
    ;
RUN;

PROC PRINT DATA=sample_data; 
RUN;


Explanation:

  • The DATA statement starts a new data step, creating a dataset named sample_data.
  • The input statement defines the variables: ID, Name (character type indicated by $), Age, Height, and Weight.
  • DATALINES allows you to enter data manually.
  • PROC PRINT prints the dataset to verify the data.

R

In R, you can use the data.frame function to create a similar dataset.

# R Example: Basic Data Frame
sample_data <- data.frame(
  ID = 1:4,
  Name = c("John", "Jane", "Jim", "Jill"),
  Age = c(25, 30, 22, 28),
  Height = c(68, 65, 70, 64),
  Weight = c(150, 120, 180, 135)
)

print(sample_data)


Explanation:

  • The data.frame function creates a data frame named sample_data.
  • Each variable (ID, Name, Age, Height, Weight) is defined with corresponding values.
  • print outputs the data frame to the console for verification.


Method 2: Using Random Number Functions

SAS

SAS provides random number functions that can be used within a DATA step.

/* SAS Example: Using Random Number Functions */
DATA random_data;
    DO ID = 1 TO 10;
        Age = ROUND(18 + 47*RAND('UNIFORM'));
        Height = ROUND(150 + 50*RAND('NORMAL'), 0.1);
        Weight = ROUND(100 + 100*RAND('NORMAL'), 0.1);
        OUTPUT;
    END;
RUN;

PROC PRINT DATA=random_data; 
RUN;


Explanation:

  • DATA starts a new data step, creating a dataset named random_data.
  • DO ID = 1 TO 10 generates 10 rows with IDs from 1 to 10.
  • RAND('UNIFORM') and RAND('NORMAL') generate random values for Age, Height, and Weight.
  • ROUND rounds the generated values.
  • OUTPUT writes the generated row to the dataset.
  • PROC PRINT prints the dataset.

R

In R, you can use functions like sample and rnorm to generate random data.

# R Example: Using Random Number Functions
set.seed(123)  # For reproducibility

random_data <- data.frame(
  ID = 1:10,
  Age = sample(18:65, 10, replace = TRUE),
  Height = rnorm(10, mean = 170, sd = 10),
  Weight = rnorm(10, mean = 70, sd = 15)
)

print(random_data)


Explanation:

  • set.seed(123) ensures reproducibility of random numbers.
  • data.frame creates a data frame named random_data.
  • sample generates random ages between 18 and 65.
  • rnorm generates normally distributed values for Height and Weight.
  • print outputs the data frame.

Method 4: Using Specialized Packages

SAS

SAS's PROC SURVEYSELECT can be used to create random samples from an existing dataset.

/* SAS Example: Using PROC SURVEYSELECT */
DATA population;
DO ID = 1 TO 1000;
Age = FLOOR(18 + RAND('UNIFORM')*47);
OUTPUT;
END;
RUN;

PROC SURVEYSELECT DATA=population OUT=sample_data SAMPSIZE=10;
RUN;

PROC PRINT DATA=sample_data;
RUN;

Explanation:

  • DATA creates a dataset named population with 1000 rows.
  • FLOOR and RAND('UNIFORM') generate random ages.
  • PROC SURVEYSELECT selects a random sample of 10 rows from population, creating sample_data.
  • PROC PRINT prints the sampled dataset.

R

In R, you can use the dplyr package for powerful data manipulation and sampling.

# R Example: Using dplyr Package
library(dplyr)

population <- tibble(
  ID = 1:1000,
  Age = floor(18 + runif(1000, min = 0, max = 47))
)

sample_data <- population %>% sample_n(10)

print(sample_data)


Explanation:

  • library(dplyr) loads the dplyr package.
  • tibble creates a tibble named population with 1000 rows.
  • floor and runif generate random ages.
  • sample_n selects a random sample of 10 rows from population.
  • print outputs the sampled data.

Method 5: Using Inline Data Entry

SAS

SAS allows inline data entry using CARDS or DATALINES.

/* SAS Example: Using CARDS or DATALINES */
DATA small_data;
    INPUT ID Name $ Age;
    DATALINES;
    1 John 25
    2 Jane 30
    3 Jim 22
    4 Jill 28
    ;
RUN;

PROC PRINT DATA=small_data; RUN;


Explanation:

  • DATA starts a new data step, creating a dataset named small_data.
  • INPUT defines the variables.
  • DATALINES allows manual data entry.
  • PROC PRINT prints the dataset.

R

In R, the tribble function from the tibble package provides a convenient way to enter data inline.

# R Example: Using tribble Function library(tibble) small_data <- tribble( ~ID, ~Name, ~Age, 1, "John", 25, 2, "Jane", 30, 3, "Jim", 22, 4, "Jill", 28 ) print(small_data)


Explanation:

  • library(tibble) loads the tibble package.
  • tribble creates a tibble named small_data.
  • Each variable is defined with corresponding values.
  • print outputs the tibble.

Summary

Both SAS and R offer a variety of methods to create sample data, each suited to different scenarios and preferences. Whether you prefer the structured environment of SAS or the flexible ecosystem of R, understanding these methods can help you efficiently generate sample datasets for your analyses.

By mastering these techniques, you can streamline your data preparation process and focus more on deriving insights from your data. Happy coding!

Feel free to leave a comment or reach out if you have any questions or need further clarifications on creating sample data in SAS or R. Stay tuned for more posts comparing these two powerful programming languages!

Popular posts from this blog

Calculating Study Day in R for CDISC Compliance: A Step-by-Step Guide

  Calculating the Study Day ( --DY ). This is a crucial step in clinical trial data analysis, and we’ll show you how to create a dynamic function in R to handle this task, even when dealing with partial dates. Function Overview The  calculate_study_day  function calculates the Study Day for clinical trial events relative to a reference date (typically the date of first dose). It handles partial dates by setting the Study Day to  NA  when dates are incomplete. Function Definition Here’s the complete function definition:   R # Load necessary package library(dplyr) # Function to calculate Study Day calculate_study_day <- function(data, subject_col, date_col, ref_date_col, study_day_col) {   temp_date_col <- paste0(date_col, "_temp")   temp_ref_date_col <- paste0(ref_date_col, "_temp")      data <- data %>%     mutate(       !!temp_date_col := as.Date(ifelse(nchar(get(date_col)) <...
Read more

Comparing SAS and R for Clinical Trial Analysis: Pros and Cons

When it comes to clinical trial data analysis, SAS and R are two of the most popular tools. Each has its own strengths and weaknesses, making them suitable for different aspects of clinical trial analysis. Let’s dive into a comparison of SAS and R to help you decide which tool might be best for your needs. Feature SAS R Cost Commercial software with significant licensing fees Open-source and free to use Regulatory Acceptance Widely accepted by regulatory bodies like FDA and EMA Gaining acceptance but not as universally recognized Data Management Robust data management capabilities Flexible but can struggle with very large datasets Statistical Procedures Comprehensive range of advanced statistical procedures Extensive statistical techniques supported by numerous packages Learning Curve Steep learning c...
Read more

Unlocking Data Consistency: Tackling Missing Variables in SDTM and ADaM Programming

In the world of clinical data programming, maintaining data integrity is absolutely essential. One of the common hurdles we face is dealing with permissible variables in SDTM (Study Data Tabulation Model) datasets that might have missing values across all records. These variables often get dropped in the final SDTM dataset but are still needed for ADaM (Analysis Data Model) datasets. This isn't just a problem for AE (Adverse Events) datasets; it extends to all SUPPxx domains and other SDTM domains as well. Why This Matters When permissible variables are missing from SDTM datasets, several issues can arise: Programming Errors : When expected variables are not present, it can cause errors in the programming code, such as variable not found errors, which can halt the data processing workflow and require additional debugging time. Data Loss : Important variables might be unintentionally omitted, leading to incomplete datasets that miss crucial information. Inconsistencies : The absence...
Read more