Numeric types (double, single, integer) in MATLAB

Introduction

Numeric types tell MATLAB how to store and use numbers. They help save memory and control precision.

When you want to save memory by using smaller number types.

When you need more precise decimal numbers for calculations.

When working with whole numbers and want faster operations.

When interfacing with hardware or files that require specific number formats.

Syntax

MATLAB

% Create variables with different numeric types
A = double(3.14);
B = single(3.14);
C = int32(10);

double is the default numeric type in MATLAB.

Use functions like double(), single(), or int32() to convert values.

Examples

This creates a double precision number, a single precision number, and a 16-bit integer.

MATLAB

x = double(5.5);
y = single(5.5);
z = int16(5);

Here, int8 stores signed 8-bit integers, and uint8 stores unsigned 8-bit integers.

MATLAB

a = int8(-100);
b = uint8(200);

Shows how single precision stores fewer decimal places than double.

MATLAB

val = single(1.23456789);
disp(val);

Sample Program

This program creates variables of different numeric types and prints their types. It also shows how double and single precision differ in decimal accuracy.

MATLAB

% Sample program showing numeric types
A = 10; % default double
B = single(10); % single precision
C = int32(10); % 32-bit integer

fprintf('A is type: %s\n', class(A));
fprintf('B is type: %s\n', class(B));
fprintf('C is type: %s\n', class(C));

% Show difference in precision
x_double = 1.123456789012345;
x_single = single(x_double);

fprintf('Double precision: %.15f\n', x_double);
fprintf('Single precision: %.15f\n', x_single);

Important Notes

Double precision uses 64 bits and is MATLAB's default for numbers.

Single precision uses 32 bits and saves memory but has less accuracy.

Integer types store whole numbers and come in different sizes like int8, int16, int32, and uint8, uint16, etc.

Summary

Numeric types control how MATLAB stores numbers in memory.

Double is default and most precise; single uses less memory but less precise.

Integer types store whole numbers and are useful for specific needs like memory saving or hardware interfacing.

Practice

1. Which numeric type is the default in MATLAB when you create a number like 5 without specifying its type?

easy

A. single

B. uint8

C. int32

D. double

Solution

1. Step 1: Understand MATLAB default numeric type

   By default, MATLAB stores numbers as double precision floating-point values.

2. Step 2: Check the given number type

   The number 5 without any type specification is stored as double.

3. Final Answer:

   double -> Option D

4. Quick Check:

   Default numeric type = double [OK]

Hint: Remember: MATLAB defaults to double unless specified [OK]

Common Mistakes:

• Thinking single is default
• Confusing integer types as default
• Assuming uint8 is default

2. Which of the following is the correct way to create a single precision number with value 3.14 in MATLAB?

easy

A. x = single(3.14);

B. x = double(3.14);

C. x = int32(3.14);

D. x = single '3.14';

Solution

1. Step 1: Identify correct syntax for single precision

   Use the function single() with a numeric argument to create a single precision number.

2. Step 2: Check each option

   x = single(3.14); uses single(3.14) which is correct syntax. x = double(3.14); creates double, not single. x = int32(3.14); creates integer, which truncates decimals. x = single '3.14'; is invalid syntax.

3. Final Answer:

   x = single(3.14); -> Option A

4. Quick Check:

   Use single() function for single precision [OK]

Hint: Use single() function to convert numbers to single [OK]

Common Mistakes:

• Using double() instead of single()
• Trying to assign string to single
• Using int32 for decimals

3. What will be the output of the following MATLAB code?

x = int8(127); y = x + 1; disp(y);

medium

A. -128

B. 128

C. Error: Overflow

D. 127

Solution

1. Step 1: Understand int8 range

   int8 stores integers from -128 to 127. Adding 1 to 127 causes overflow.

2. Step 2: Check overflow behavior

   In MATLAB, int8 overflow wraps around to the minimum value, so 127 + 1 becomes -128.

3. Final Answer:

   -128 -> Option A

4. Quick Check:

   int8 overflow wraps to -128 [OK]

Hint: int8 max is 127; adding 1 wraps to -128 [OK]

Common Mistakes:

• Expecting 128 (out of range)
• Expecting error on overflow
• Thinking value stays at 127

4. Consider this MATLAB code:

a = single(1.5); b = int16(a); disp(b);

What is the output and why?

medium

A. 1.5

B. 1

C. 2

D. Error: Cannot convert single to int16

Solution

1. Step 1: Convert single to int16 behavior

   When converting from single to int16, MATLAB truncates the decimal part (does not round).

2. Step 2: Apply truncation to 1.5

   Truncating 1.5 results in 1, so b becomes 1.

3. Final Answer:

   1 -> Option B

4. Quick Check:

   Conversion truncates decimals, so 1.5 -> 1 [OK]

Hint: Conversion truncates decimals, no rounding [OK]

Common Mistakes:

• Expecting rounding to 2
• Expecting error on conversion
• Expecting output as 1.5

5. You want to store a large array of numbers between 0 and 255 efficiently in MATLAB. Which numeric type should you choose to minimize memory while preserving all values exactly?

hard

A. double

B. single

C. uint8

D. int16

Solution

1. Step 1: Identify value range and memory needs

   Values are between 0 and 255, so an unsigned 8-bit integer (uint8) fits exactly.

2. Step 2: Compare types for memory and precision

   double and single use more memory and are floating-point. int16 uses 16 bits and allows negative values, wasting space.

3. Step 3: Choose the best type

   uint8 uses 8 bits, stores 0-255 exactly, and saves memory.

4. Final Answer:

   uint8 -> Option C

5. Quick Check:

   Use uint8 for 0-255 integer range to save memory [OK]

Hint: Use uint8 for 0-255 integers to save memory [OK]

Common Mistakes:

• Choosing double wastes memory
• Using int16 wastes space and allows negatives
• Using single loses integer precision