The primitive data types are the most fundamental types in Kotlin; all other types are built up of these types and arrays thereof. Their representation is very efficient (both in terms of memory and CPU time), as they map to small byte groups that are directly manipulatable by the CPU.
Integer types in Kotlin have a limited size, as opposed to the arbitrarily large integers in Python. The limit depends on the type, which decides how many bits the number occupies in memory:
|Type||Bits||Min value||Max value|
Bytes are -128 through 127 due to Kotlin inheriting a bad design decision from Java. In order to get a traditional byte value between 0 and 255, keep the value as-is if it is positive, and add 256 if it is negative (so -128 is really 128, and -1 is really 255). See the section on extension functions for a neat workaround for this.
An integer literal has the type
Int if its value fits in an
Long literals should be suffixed by
L for clarity, which will also let you make a
Long with a "small" value. There are no literal suffixes for
Byte, so such values need an explicit type declaration or the use of an explicit conversion function.
val anInt = 3 val anotherInt = 2147483647 val aLong = 2147483648 val aBetterLong = 2147483649L val aSmallLong = 3L val aShort: Short = 32767 val anotherShort = 1024.toShort() val aByte: Byte = 65 val anotherByte = -32.toByte()
Beware that dividing an integer by an integer produces an integer (like in Python 2, but unlike Python 3). If you want a floating-point result, at least one of the operands needs to be a floating-point number (and recall that like in most languages, floating-point operations are generally imprecise):
println(7 / 3) // Prints 2 println(7 / 3.0) // Prints 2.3333333333333335 val x = 3 println(7 / x) // Prints 2 println(7 / x.toDouble()) // Prints 2.3333333333333335
Whenever you use an arithmetic operator on two integers of the same type (or when you use a unary operator like negation), there is no automatic "upgrading" if the result doesn't fit in the type of the operands! Try this:
val mostPositive = 2147483647 val mostNegative = -2147483648 println(mostPositive + 1) println(-mostNegative)
Both of these print
-2147483648, because only the lower 32 bits of the "real" result are stored.
When you use an arithmetic operator on two integers of different types, the result is "upgraded" to the widest type. Note that the result might still overflow.
In short: think carefully through your declarations of integers, and be absolutely certain that the value will never ever need to be larger than the limits of the type! If you need an integer of unlimited size, use the non-primitive type
| ||64||16-17 significant digits (same as |
| ||32||6-7 significant digits|
| ||16||UTF-16 code unit (see the section on strings - in most cases, this is one Unicode character, but it might be just one half of a Unicode character)|
Floating-point numbers act similarly to in Python, but they come in two types, depending on how many digits you need. If you need larger precision, or to work with monetary amounts (or other situations where you must have exact results), use the non-primitive type
|← Previous: Declaring variables||Next: Strings →|
This material was written by Aasmund Eldhuset; it is owned by Khan Academy and is licensed for use under CC BY-NC-SA 3.0 US. Please note that this is not a part of Khan Academy's official product offering.
© 2010–2020 JetBrains s.r.o. and Kotlin Programming Language contributors
Licensed under the Apache License, Version 2.0.