# Java data types: conversion

Jason Attin

Ranch Hand

Posts: 232

2

posted 1 month ago

Hi guys, this question on Glenn, Mitchell. OCAJP Oracle Certified Associate Java SE 8 Programmer Practice Exams (Kindle Locations 19486-19488). Enthuware. Kindle Edition.

I got this wrong and saiid it prints 0, but it in fact prints -46. Now how am I supposed to know that without compiling it and run it? Is there a general rule I can get out of this? The explanation says

Is that the case for only numbers with 9 digits?

Consider the following code:

What will it print when run? Select 1 option

A. It will print 0.

B. It will not print 0.

C. It will not compile.

D. It will throw an exception at runtime.

E. None of the above.

I got this wrong and saiid it prints 0, but it in fact prints -46. Now how am I supposed to know that without compiling it and run it? Is there a general rule I can get out of this? The explanation says

the information was lost during the conversion from type int to type float as values of type float are not precise to nine significant digits.

Is that the case for only numbers with 9 digits?

posted 1 month ago

I would really be careful with this -- as it is more like "around 9 digits".

First, notice that phrase "significant". So, how many significant digits does "12340000000" have?

Second, while you understand 9 digits in scientific notation, the computer doesn't store it that way. Floating point is stored as a binary number that will be multiplied by 2 to the power of another binary number. And there may be lots of precision loss during the conversion of a decimal number that will be multiplied by 10 to the power of another decimal number. Add to that, normalization, blah, blah, blah ...

It is more like "around 9 significant digits". It's a rough estimate. So, don't depend on it.

Henry

Jason Attin wrote: The explanation saysthe information was lost during the conversion from type int to type float as values of type float are not precise to nine significant digits.

Is that the case for only numbers with 9 digits?

I would really be careful with this -- as it is more like "around 9 digits".

First, notice that phrase "significant". So, how many significant digits does "12340000000" have?

Second, while you understand 9 digits in scientific notation, the computer doesn't store it that way. Floating point is stored as a binary number that will be multiplied by 2 to the power of another binary number. And there may be lots of precision loss during the conversion of a decimal number that will be multiplied by 10 to the power of another decimal number. Add to that, normalization, blah, blah, blah ...

It is more like "around 9 significant digits". It's a rough estimate. So, don't depend on it.

Henry

It is sorta covered in the JavaRanch Style Guide. |