Why does a float fit in a double but a float[] doesn't fit a double[]?

Question

If I have a function that takes a double as an argument, I can easily put in a float.

However when I have a funcion that takes a double[] then I cannot pass a float[].

public static void doSomethingWithMyDoubles(double[] doubles) {
    //....
}

public static void doSomethingWithMyDouble(double doubles) {
    //....
}

public static void main() {
    float floater = 10f;
    double doubler = 10d;

    doSomethingWithMyDouble(doubler) // OK
    doSomethingWithMyDouble(floater) // OK

    float[] floates = new float[10];
    double[] doubles = new double[10];

    doSomethingWithMyDoubles(doubles) // OK
    doSomethingWithMyDoubles(floates) // NOK
}

The best is to forget about `float` and behave as if there was only `double`. — Ingo, Mar 07 '14 at 14:31
Lol! If only there were no such things/people as embedded software/programmers :) — Timo, Mar 07 '14 at 14:32
I didn't say I was working on a car, but it might depending on the type and age I guess. — Timo, Mar 07 '14 at 14:56

score 2 · Accepted Answer · edited May 23 '17 at 10:31

2

In Java, arrays are not a primitive data type; instead they are of type Object. As a consequence, you cannot convert an array of one primitive type to an array of another primitive type directly (even though you might cast the individual elements in the array to the other type).

edited May 23 '17 at 10:31

Community

1
1

answered Mar 07 '14 at 14:27

Elliott Frisch

198,278
20
158
249

Well, not for primitives. A `String[]` can still be used as an `Object[]` for example... – Jon Skeet Mar 07 '14 at 14:29
Not entirely true, because you can cast an Object[] to a String[], just like you can cast an Object to a String. But you cannot cast a float[] to a double[] like you can cast a float to a double. – Erwin Bolwidt Mar 07 '14 at 14:30
@JonSkeet Edited for clarity. Thanks! – Elliott Frisch Mar 07 '14 at 14:32

score 2 · Answer 2 · answered Mar 07 '14 at 14:30

2

It is because implicite type conversion takes place when you pass float value into the method whitch requires double one. But there is no type conversion between float[] and double[] arrays.

answered Mar 07 '14 at 14:30

Megaprog

466
2
13

score 2 · Answer 3 · answered Mar 07 '14 at 14:36

A float is a 4-byte type, and a double is an 8-byte type. Widening a single float to a double is a constant-time operation. However, widening a float[] to a double[] has to do a full copy, and needs to take linear time to widen the entire array. So Java won't do it implicitly -- it won't pretend that the operation is as cheap as an O(1) operation.

score 0 · Answer 4 · answered Mar 07 '14 at 14:29

0

That is because float and doubles are primitives. Floats can be automatically casted into doubles. However, double[] and float[] are objects. Try explicitly casting. You may need to iterate and cast each element and insert into your new array.

answered Mar 07 '14 at 14:29

Brian

7,098
15
56
73

Why does a float fit in a double but a float[] doesn't fit a double[]?

4 Answers4