should we use float as primary key in sql-server

Question

some one suggested me its a better idea to use float as a primary key of a table instead of using the BIGINT. can we make float primary key to be identity ?

Answer 1

Considerations:

• you CAN make a float field primary key.
• you CANT make a float field IDENTITY. Identity columns must be of data type int, bigint, smallint, tinyint, or decimal or numeric with a scale of 0
• you ABSOLUTELY SHOULD NOT use a float field as a PK. Again, you CAN, SQL Server will allow you but its not recommended mainly because floats are inaccurate as @Andy said.

why do you need a float as a PK anyway? Do you need a value like 3,1234235234534 to uniquely identify your row?

Answer 2

Notice how if you do float a=1f and float b=1f they are the same right ?

However if(a==b) will likely not be true as floats are inaccurate.