Numbers get stored at some place in memory. They have a certain memory size (spaces in memory) and an arrangement for their bytes.
Digital numbers
The data stored in memory is just a lot of tiny electronic connections in the circuits of the micro:bit. A number is the most basic type of data and other types of data are made using numbers (number are even used to act as text characters in memory). A number is made with combination of electronic connections that are either on and off. One connection is called a bit and this is used as a single digit for computer numbers. A single bit can mean one of two numbers, 0 or 1. Two bits can mean one of four numbers, 00, 01, 10, or 11. We use decimal numbers normally so we give the decimal names to the computer numbers. So, for the two bit numbers, their decimal names are: 00 = 0, 01 = 1, 10 = 2, and 11 = 3. When you see numbers with ones and zeros, they are called binary numbers.
The ‘byte’
The smallest amount of bits normally used to store a number in a computer is a combination of 8 bits. A group of 8 bits is called a byte. A byte looks like 10010010. This binary number, for example, represents the number 162 in decimal.
The maximum absolute value for a number contained in a byte is 255. You couldn’t store the decimal number of 2552 in one byte, you need to use two bytes together. So, 2552 as a binary number is 00001001 11111000 using two bytes.
Numbers are stored in memory in different ways. They can use one or more bytes, have a positive or negative value, and have their bytes switched around.
Signed numbers
If you want a number that can have a positive value or a negative value, then you use a signed number. This means that one bit of all the bits than make up the number value is treated as a plus sign or a minus sign. A bit value of 0 means a plus and a bit value of 1 means minus. Using one of the bits of the number for a sign means that the maximum possible value for a number gets reduced by about half.
As an example, the decimal numbers of 1 and -1 are 00000001 and 11111111 in binary. You see that the bit on the left of the second binary number is a 1. That bit is used as the minus sign (-) for the -1.
Two’s complement
You might think that the binary number for a decimal -1 would be 10000001 if the left bit is a 1 for the minus sign. The binary number for a decimal -1 is actually 11111111. That’s because computers and electronic devices use a special trick when working with negative numbers. The trick is called two’s complement.
Making a negative number from a positive number is a two step process. The first step is the one’s complement of the number. This step switches all the bits in the number that are 1 to 0 and all the bits that are 0 to 1 (invert the bits). Then, the twos’s complement step adds the value of 1 to the one’s complement value. Here’s how it works:
Start with the positive binary number for a decimal 1 which is 00000001.
- The one’s complement switches the bits, the binary number is now
11111110.
- The two’s complement adds a binary
1 to the one’s complement value.
11111110 + 00000001 = 11111111
Unsigned numbers
Signed numbers use all of their bits for the value itself and are always positive values.
Big end and little end (endian)
Earlier you saw that the decimal number 2552 needs two bytes in memory. The order in which these two bytes are placed in memory is called endian order or endianness. Funny word, right? This comes from the idea that the byte with the larger part of the value is called the big end and the byte with smaller part of the value is called the little end.
For 2552 its binary number uses two bytes which are: 00001001 11111000. The two parts (ends) of this number are:
- Big end:
00001001
- Little end:
11111000
If the big end of the number is stored in memory first, before the little end, the number is a big endian number. Then, of course, if the little end of the number is stored in memory first, before the big end, the number is a little endian number.
Sometimes you need to have your program tell what type of numbers it will store in memory. This often necessary when you use pin operations with a buffer.
The formats for numbers stored on the micro:bit are:
Int8LE: one byte, signed, little endian UInt8LE: one byte, unsigned, litte endianInt8BE: one byte, signed, big endianUInt8BE: one byte, unsigned, big endianInt16LE: two bytes, signed, little endianUInt16LE: two bytes, unsigned, little endianInt16BE: two bytes, signed, big endianUInt16BE: two bytes, unsigned, big endianInt32LE: four bytes, signed, little endianInt32BE: four bytes, signed, big endian
See also
buffer
