← Back to Overview

Data Types and Their Sizes

Concept Explanation

What is it?

Data types classify the kind of values that a variable can hold (e.g., numbers, text, true/false). Understanding the "size" of a data type refers to the amount of memory allocated to store a value of that type. This can be fixed (as in C/Java primitives) or dynamic (as in Python, or Oracle's NUMBER type).

Why is it important?

Knowing data type sizes is crucial for efficient memory management, preventing overflow errors, optimizing performance, and understanding how data is represented in memory. It directly impacts application performance and resource consumption.

Where is it used?

Memory-constrained environments (embedded systems, game development), performance-critical applications, low-level programming (C), and database design (Oracle) all require careful consideration of data types and their storage.

Real-world example

Imagine storing the age of a person (usually a small integer) versus storing a person's name (a variable-length string). Using a fixed-size small integer type for age is efficient, while a flexible string type is necessary for names to accommodate varying lengths. Knowing the size limits helps prevent names from being cut off or wasting memory.

---

Algorithm

1. Start.
2. For each supported data type: a. Determine or retrieve its size/memory usage. b. Display the data type and its corresponding size.
3. End.

Edge Cases:

• Languages with dynamic typing where "size" is object-dependent.
• Database types where storage is more about precision/scale than fixed byte size.
• Platform-dependent sizes (e.g., long in C).

---

Implementations

Java (JDK 17)

public class DataTypeSizes {
    public static void main(String[] args) {
        System.out.println("Size of byte: " + Byte.BYTES + " bytes (" + Byte.SIZE + " bits)");
        System.out.println("Size of short: " + Short.BYTES + " bytes (" + Short.SIZE + " bits)");
        System.out.println("Size of int: " + Integer.BYTES + " bytes (" + Integer.SIZE + " bits)");
        System.out.println("Size of long: " + Long.BYTES + " bytes (" + Long.SIZE + " bits)");
        System.out.println("Size of float: " + Float.BYTES + " bytes (" + Float.SIZE + " bits)");
        System.out.println("Size of double: " + Double.BYTES + " bytes (" + Double.SIZE + " bits)");
        // Boolean and char do not have a .BYTES constant
        System.out.println("Size of char: " + Character.SIZE + " bits");
        // Boolean size is JVM dependent, often 1 byte for arrays, but not precisely defined for a single variable
        System.out.println("Size of boolean: JVM dependent, usually 1 byte for arrays");
    }
}

Python (3.10+)

import sys

# Example values for different data types
int_val = 0
large_int_val = 1234567890123456789012345678901234567890
float_val = 3.14159
str_val_empty = ""
str_val_hello = "Hello"
list_val_empty = []
list_val_nums = [1, 2, 3]
tuple_val_empty = ()
tuple_val_nums = (1, 2, 3)
dict_val_empty = {}
dict_val_items = {'a': 1, 'b': 2}
bool_true = True
bool_false = False
none_val = None

print(f"Size of int (0): {sys.getsizeof(int_val)} bytes")
print(f"Size of large int: {sys.getsizeof(large_int_val)} bytes") # Size increases with value
print(f"Size of float (3.14): {sys.getsizeof(float_val)} bytes")
print(f"Size of empty string ('{str_val_empty}'): {sys.getsizeof(str_val_empty)} bytes")
print(f"Size of string ('{str_val_hello}'): {sys.getsizeof(str_val_hello)} bytes") # Size increases with characters
print(f"Size of empty list ([]): {sys.getsizeof(list_val_empty)} bytes")
print(f"Size of list ([1,2,3]): {sys.getsizeof(list_val_nums)} bytes") # Size increases with elements
print(f"Size of empty tuple (()): {sys.getsizeof(tuple_val_empty)} bytes")
print(f"Size of tuple ((1,2,3)): {sys.getsizeof(tuple_val_nums)} bytes")
print(f"Size of empty dict ({{}}): {sys.getsizeof(dict_val_empty)} bytes")
print(f"Size of dict ({{'a':1, 'b':2}}): {sys.getsizeof(dict_val_items)} bytes")
print(f"Size of boolean (True): {sys.getsizeof(bool_true)} bytes")
print(f"Size of boolean (False): {sys.getsizeof(bool_false)} bytes")
print(f"Size of None: {sys.getsizeof(none_val)} bytes")

# Note: The sizes reported by sys.getsizeof() include the object's overhead
# and can vary slightly between Python versions and system architectures.
# For strings and collections, the size also depends on the content and number of elements.

C (C99)

#include <stdio.h>

int main() {
    printf("Size of char: %zu byte(s)\n", sizeof(char));
    printf("Size of short: %zu byte(s)\n", sizeof(short));
    printf("Size of int: %zu byte(s)\n", sizeof(int));
    printf("Size of long: %zu byte(s)\n", sizeof(long));
    printf("Size of long long: %zu byte(s)\n", sizeof(long long));
    printf("Size of float: %zu byte(s)\n", sizeof(float));
    printf("Size of double: %zu byte(s)\n", sizeof(double));
    printf("Size of long double: %zu byte(s)\n", sizeof(long double));
    printf("Size of void*: %zu byte(s)\n", sizeof(void*)); // Size of a pointer

    return 0;
}

Oracle (19c)

-- In Oracle, data types like NUMBER and VARCHAR2 are dynamically sized or have flexible storage.
-- Their "size" refers more to their maximum capacity and precision than a fixed byte allocation.
SET SERVEROUTPUT ON;
BEGIN
  DBMS_OUTPUT.PUT_LINE('Oracle Data Type Storage Characteristics (not fixed byte size like C/Java primitives)');
  DBMS_OUTPUT.PUT_LINE('---------------------------------------------------------------------------------');
  DBMS_OUTPUT.PUT_LINE('NUMBER: Variable (up to 38 digits precision).');
  DBMS_OUTPUT.PUT_LINE('VARCHAR2(N): Variable (up to N bytes/chars), max 4000 in SQL, 32767 in PL/SQL.');
  DBMS_OUTPUT.PUT_LINE('CHAR(N): Fixed N bytes/chars, max 2000.');
  DBMS_OUTPUT.PUT_LINE('DATE: Fixed 7 bytes.');
  DBMS_OUTPUT.PUT_LINE('TIMESTAMP: Fixed 7 or 11 bytes.');
  DBMS_OUTPUT.PUT_LINE('BOOLEAN (PL/SQL only): Internally handled, not a SQL type.');
  DBMS_OUTPUT.PUT_LINE('---------------------------------------------------------------------------------');
  DBMS_OUTPUT.PUT_LINE('Note: Actual storage can vary based on value and database configuration.');
END;
/

Explanation

• Java: Primitive types (byte, short, int, long, float, double, boolean, char) have fixed sizes across all JVMs, ensuring platform independence. Wrapper classes provide BYTES and SIZE constants.
• Python: Dynamically typed. The "size" refers to the memory an object consumes, which can vary based on its value and content. sys.getsizeof() provides this, but includes object overhead.
• C: Data type sizes can vary by compiler and system architecture (e.g., 32-bit vs. 64-bit systems). The sizeof operator returns the size in bytes. Pointers also have a size that matches the system's address bus width.
• Oracle: Data types like NUMBER and VARCHAR2 have flexible storage. NUMBER storage depends on precision, and VARCHAR2 on the actual string length. DATE and TIMESTAMP have fixed, small storage. BOOLEAN is a PL/SQL-only type.

Complexity Analysis

• Time Complexity: O(1) - The operation of querying data type sizes is constant time.
• Space Complexity: O(1) - Minimal extra space is used for storing the example data types and printing their sizes.

Flowchart

Sample Dry Run

Since this program primarily outputs fixed sizes (or size characteristics), a traditional step-by-step dry run with changing input isn't directly applicable. The output will consistently show the byte sizes or characteristics as demonstrated in the implementation section.

Practice Problems

Easy

• Modify the program to display the minimum and maximum values that each data type can hold.
• For C, try compiling and running on different architectures (if possible) to observe sizeof variations.

Medium

• For Python, explore how sys.getsizeof() changes for lists and dictionaries as you add elements.
• In Oracle, declare variables with different NUMBER precisions (e.g., NUMBER(5), NUMBER(10,2)) and discuss their implications.

Hard

• Investigate how data alignment affects the memory footprint of structs in C or objects in Java.
• Research how memory is managed for different data types in the runtime environments (JVM, C runtime, Python interpreter, Oracle database).

---

"Measuring programming progress by lines of code is like measuring aircraft building progress by weight." - Bill Gates