February 15, 2023•4 min read

Getting to Know the Float Data Type in Java

Float, or floating point, is a data type that holds decimal and fractional number literals.

Read in Indonesian

After getting to know the integer data type, this time we'll look at another numeric data type, float.

About the Float Data Type

Floating Points, or usually float, is a standard data type that holds decimal numbers. If integer only holds whole numbers, float also holds fractional ones. For example: 1.5, 2.2, 3.14, and so on.

Float Sub Types

This data type splits into 2 based on precision — how many digits sit after the dot, namely:

  • float: a 32-bit (4 byte) data type that holds 6 to 7 digits after the dot. The primitive type is float, while the object form is Float.
  • double: a 64-bit (8 byte) data type that holds 15 to 16 digits after the dot. The primitive type is double, while the object form is Double.

As you can see, the higher the precision, the bigger the size. Since computers today are more capable, using double doesn't make a significant difference compared with float. Still, saving memory by using float whenever possible is a good idea.

Float Literals

A float literal is any real number. That covers positive, negative, whole and fractional numbers, and zero.

Unlike decimal numbers in Indonesia, which are separated by a comma (,), decimal numbers in Java use a dot (.).

For example:

JAVA
double PI = 3.141592653589793; // 15 digits after the dot

float  tiga  = 3;   // whole number
double empat = 4.0; // whole number

Using Letter Suffixes

You can also add the letter f or d to mark a number as float or double. This is strongly recommended as a type marker for a literal.

For example:

JAVA
float  pi = 3.141592f;          // the `f` suffix is required
double PI = 3.141592653589793d; // the `d` suffix is optional

float  tiga  = 3f;   // the `f` suffix is optional
float  empat = 4.0f; // the `f` suffix is required
double lima  = 5d;   // the `d` suffix is optional

By default Java reads decimal numbers as the double data type, which is why the f suffix is needed to mark a literal as float. Without it the program errors out, because Java can't convert a double to a float automatically.

When doing arithmetic, division especially, this marker matters a lot because it can change the result. We'll cover that further in the next section.

Using Scientific Notation

A float literal can also be written in scientific notation. In case you forgot, scientific notation uses multiplication by powers of 10. More on that here.

Java uses the e notation to write the literal. For example like this:

JAVA
double seribu  = 1e3;  // 1 * 10^3 = 1000
float  duaJuta = 2e6f; // 2 * 10^6 = 2000000

double bilanganKecil  = 7e-10; // 0.0000000007
float  bilanganKecil2 = 4e-5f; // 0.00004

Operations on the Float Data Type

Every operation that works on integer works on float too. There's arithmetic, arithmetic assignment, and the unary operator. More on that in the earlier discussion.

The division operation is probably what makes it a little different. As we touched on earlier, adding the letter f or d can change the result.

Let's look at an example.

JAVA
public class TipeDataFloat{
  public static void main(String[] args) {
    float a = 22;
    float b = 7;
    float hasil = a / b;

    System.out.printf("hasil dari %.2f / %.2f adalah %.2f", a, b, hasil);
  }
}

Careful! the letter f above isn't part of a float literal, it's part of the string template for printing a float. More on that here

If you run it, the result looks like this:

TEXT
hasil dari 22.00 / 7.00 adalah 3.14

So far so good.

Now let's fill the hasil variable with an integer literal division. We'll change the code above to this:

JAVA
public class TipeDataFloat{
  public static void main(String[] args) {
    float hasil = 22 / 7;

    System.out.printf("hasil dari 22 / 7 adalah %.2f", hasil);
  }
}

Now run that code and look at the result. The calculation surely comes out wrong. Instead of printing 3.14 it prints 3.00.

TEXT
hasil dari 22 / 7 adalah 3.00

So why does that happen?

Because we use integer literals for the division. Remember, division on integers is always rounded and can never produce a decimal number.

As a result, the integer division of 22 and 7, which gives the whole number 3, goes into the hasil variable. So what lands in hasil is purely the integer result, regardless of the variable's data type.

To fix this we can use a float literal by adding the letter f after the number. So the code becomes like this:

JAVA
public class TipeDataFloat{
  public static void main(String[] args) {
    float hasil = 22f / 7f;
    // you can also write `22f / 7` or `22 / 7f`
    // but for consistency, put `f` on both

    System.out.printf("hasil dari 10 / 8 adalah %.2f", hasil);
  }
}

Next Up

Now that the 2 numeric data types in Java are done (integer and float), next we'll move to data types that hold letters and characters.

Next we'll talk about the 'char' data type in the Java language.

Hope it helps, happy coding 👋

Get better at shipping code.

Join the other developers getting tech updates, tips, and relaxed tutorials in their inbox every week.

Stay in the loop

Tech updates, coding tips, and relaxed insights straight to your inbox every week.

No spam. Unsubscribe anytime.

Read next

See all

For developers who learn by struggling

© 2026 Inva.dev.