# How is -Integer.MIN_VALUE Equal to Integer.MIN_VALUE

Avinash Haridasu

Ranch Hand

Posts: 31

Campbell Ritchie

Marshal

Posts: 52531

119

posted 5 years ago

Write out its bit pattern. HInt: you can get it from Integer.toBinaryString(Integer.Min_VALUE). It might be 1 and thirty-one 0s. Or it might not be.

Work out its two's complement when negated: Hint: Write out the bit pattern for 2 to the 32nd. That is 1 and thirty-two 0s. Subtract the binary pattern for -2147483648 from that. And what do you get . . . Tadaaa!

Google for arithmetic overflow and two's complement binary integers, and you will get a different explanation of the same thing.

Work out its two's complement when negated: Hint: Write out the bit pattern for 2 to the 32nd. That is 1 and thirty-two 0s. Subtract the binary pattern for -2147483648 from that. And what do you get . . . Tadaaa!

Google for arithmetic overflow and two's complement binary integers, and you will get a different explanation of the same thing.

Campbell Ritchie

Marshal

Posts: 52531

119

Avinash Haridasu

Ranch Hand

Posts: 31

Rameshwar Soni

Ranch Hand

Posts: 247

posted 5 years ago

.

In binary number system we can use only 2 things and that is 0 and 1 and we have to use 0 and 1 for representing the negative numbers also.There is something known as MSB(most significant bit) and LSB(least significant bit). Now the left-most-bit is your MSB and right-most-bit is LSB. Example 01010101 Here the one in red color is MSB and the one in green color is LSB..

.

So you must have understood what is MSB and LSB. Now if the MSB is 0 then the number is positive and if the MSB is 1 then the number is negative. So the conclusion is MSB represents the sign of the number and other numbers(i would say bits ) are the actual value. So

Since MSB is 0 therefore number is positive and waht about the value 0000001 is equal to 1 and therefor ans is . Now Now MSB is 1 therefore number is negative and Java uses 2's complement for negative numbers and the 2's complement of 11111111 is 00000001 and therefore the ans is -1 (Please find out how to find 2's complement if you are not aware).

.

As for your second question about >> i am not sure.

- 1

Avinash Haridasu wrote:Can any one tell me

how do we decide whether a number is negative or positive just by looking at the binary representation

0 0 0 0 0 0 0 1 = 1

1 1 1 1 1 1 1 1 = −1

.

In binary number system we can use only 2 things and that is 0 and 1 and we have to use 0 and 1 for representing the negative numbers also.There is something known as MSB(most significant bit) and LSB(least significant bit). Now the left-most-bit is your MSB and right-most-bit is LSB. Example 01010101 Here the one in red color is MSB and the one in green color is LSB..

.

So you must have understood what is MSB and LSB. Now if the MSB is 0 then the number is positive and if the MSB is 1 then the number is negative. So the conclusion is MSB represents the sign of the number and other numbers(i would say bits ) are the actual value. So

Since MSB is 0 therefore number is positive and waht about the value 0000001 is equal to 1 and therefor ans is . Now Now MSB is 1 therefore number is negative and Java uses 2's complement for negative numbers and the 2's complement of 11111111 is 00000001 and therefore the ans is -1 (Please find out how to find 2's complement if you are not aware).

.

As for your second question about >> i am not sure.

Campbell Ritchie

Marshal

Posts: 52531

119

Campbell Ritchie

Marshal

Posts: 52531

119

posted 5 years ago

Look in the Java Language Specification and you find that the number of bits shifted is calculated % i where i is the number of bits in the datatype. So for everything except a

i >> 5 is equivalent to i / 32

`long`, shift 37 and shift 5 have the same effect.i >> 5 is equivalent to i / 32

Rameshwar Soni

Ranch Hand

Posts: 247

Stephan van Hulst

Bartender

Posts: 6583

84

Campbell Ritchie

Marshal

Posts: 52531

119

Campbell Ritchie

Marshal

Posts: 52531

119

posted 5 years ago

No, Jesper actually means 6553Stephan van Hulst wrote:Did you mean -1?

__5__, assuming that is the bit pattern for a

`char`. That is equivalent to the Unicode character for ffff, and you can find what that is here.

Stephan van Hulst

Bartender

Posts: 6583

84

Campbell Ritchie

Marshal

Posts: 52531

119

Avinash Haridasu

Ranch Hand

Posts: 31

posted 5 years ago

As per Specification:

If the promoted type of the left-hand operand is int, only the five lowest-order bits of the right-hand operand are used as the shift distance. It is as if the right-hand operand were subjected to a bitwise logical AND operator & with the mask value 0x1f. The shift distance actually used is therefore always in the range 0 to 31, inclusive.

But why it is "only the five lowest-order bits" ???

Can any one please explain ??

Campbell Ritchie wrote:Look in the Java Language Specification and you find that the number of bits shifted is calculated % i where i is the number of bits in the datatype. So for everything except along, shift 37 and shift 5 have the same effect.

i >> 5 is equivalent to i / 32

As per Specification:

If the promoted type of the left-hand operand is int, only the five lowest-order bits of the right-hand operand are used as the shift distance. It is as if the right-hand operand were subjected to a bitwise logical AND operator & with the mask value 0x1f. The shift distance actually used is therefore always in the range 0 to 31, inclusive.

But why it is "only the five lowest-order bits" ???

Can any one please explain ??

Stephan van Hulst

Bartender

Posts: 6583

84

posted 5 years ago

Because you don't need any more than that. With 5 bits you can create the values 0-31, enough to perform any possible shift distance on an int.

*The mind is a strange and wonderful thing. I'm not sure that it will ever be able to figure itself out, everything else, maybe. From the atom to the universe, everything, except itself.*

Avinash Haridasu

Ranch Hand

Posts: 31

posted 5 years ago

Need small clarification on shifting:

32 In Binary: 0000 0000 - 0000 0000 - 0000 0000 - 0010 0000

32>>1:

0000 0000 - 0000 0000 - 0000 0000 - 0001 0000

32>>2:

0000 0000 - 0000 0000 - 0000 0000 - 0000 1000

32>>1:

0000 0000 - 0000 0000 - 0000 0000 - 0000 0001

32>>6:

0000 0000 - 0000 0000 - 0000 0000 - 0000 0000

When 32>>6 is itself 0 - Then how 32>>33 is 16, 32>>38 is 0 ???

32 In Binary: 0000 0000 - 0000 0000 - 0000 0000 - 0010 0000

32>>1:

0000 0000 - 0000 0000 - 0000 0000 - 0001 0000

32>>2:

0000 0000 - 0000 0000 - 0000 0000 - 0000 1000

32>>1:

0000 0000 - 0000 0000 - 0000 0000 - 0000 0001

32>>6:

0000 0000 - 0000 0000 - 0000 0000 - 0000 0000

When 32>>6 is itself 0 - Then how 32>>33 is 16, 32>>38 is 0 ???

Campbell Ritchie

Marshal

Posts: 52531

119

posted 5 years ago

The Java Language Specification link I gave you yesterday explains it all. There are 32 bits to an

32 bits in an

32 = 2 to the 5th.

So you shift an

`int`, so you cannot shift an`int`more than 31 bits. You can fit 31 into 5 bits. [A`long`can be shifted up to 63 bits, and you can fit 63 into 6 bits.] If you shifted it 32 bits, you would come back to where you started, so`i >> 32`is the same as`i`, which is the same as`i >> 0`. So you can shift an`int`by`|n % 32|`bits, where || is the absolute value operator. Now the absolute value of`n % 32`is the same (in binary arithmetic) as`n & 32 - 1`. And as we all know, 32 - 1 = 0x1f. If you write the bit pattern for 0x1f you find it has 5 1s in. So you get 5 bits masked.32 bits in an

`int`.32 = 2 to the 5th.

So you shift an

`int`by the rightmost (least significant)__five__bits (six for a`long`).