Java reads Double as a String?

Question

In this line:

totalCost += df.format(pCost);

I get the error 'inconvertable types' required: double, found: string, yet the pCost variable is a double, and the method that made it outputs it as a double.. does anyone know why this may happen?

I need it to be to 2 decimal places. I don't get why this happens at all though. — Chillo, Nov 30 '13 at 14:26
@Chillo you don't round while adding, you round when you have the final value. — Lesleh, Nov 30 '13 at 14:27
Yes, `df` is a `DecimalFormat`, which produces a `String`. "Two decimal places" doesn't have any meaning with binary floating point. — chrylis -cautiouslyoptimistic-, Nov 30 '13 at 14:27
If you're trying to round because you're dealing with money, don't ever use floating-point for money; use a money library (Joda Money) or an integral type of the cents (or other base unit). — chrylis -cautiouslyoptimistic-, Nov 30 '13 at 14:28
That is the final value, but the unrounded version is used elsewhere furthur along. — Chillo, Nov 30 '13 at 14:29

score 2 · Accepted Answer · edited May 23 '17 at 12:05

2

use just this:

totalCost += pCost;

or, if you will:

totalCost += Double.valueOf(df.format(pCost));

see: Round a double to 2 decimal places

edit: as a response to the caveat below: This code isn't truely precise; follow the link for a more accurate version (in short: convert it to a BigDecimal and use .setScale())

edited May 23 '17 at 12:05

Community

1
1

answered Nov 30 '13 at 14:26

ljgw

2,751
1
20
39

I need it to be to 2 dp, hence the df.format – Chillo Nov 30 '13 at 14:28
Saw your comment, and adjusted (and readjusted) my code. – ljgw Nov 30 '13 at 14:31
Thanks for the edit, followed that new advice and it seems to have worked! Thanks. – Chillo Nov 30 '13 at 14:32
This code isn't really accurate; even assuming that all of the doubles in question are within a close range, the "rounding" is still subject to re-rounding when the decimal is stuffed back into binary representation; the value `0.10` can't be exactly represented as a binary fraction. – chrylis -cautiouslyoptimistic- Nov 30 '13 at 14:33
3

@Chillo *Seems* to have worked. *Do not* use floating-point for money in production code. – chrylis -cautiouslyoptimistic- Nov 30 '13 at 14:33
I know. See the provided link for a more accurate function to round a double to 2 decimal places. – ljgw Nov 30 '13 at 14:34
Seen and bookmarked, ta. This is no where near production code but cheers for the pointer. – Chillo Nov 30 '13 at 14:40

score 1 · Answer 2 · answered Nov 30 '13 at 14:28

1

Well Decimal format really does return String:

DecimalFormat df = new DecimalFormat("0.00##");
String result = df.format(34.4959);

but to get a Double you could do:

Double.valueOf(df.format(34.4959));

answered Nov 30 '13 at 14:28

Kuba Spatny

26,618
9
40
63

Java reads Double as a String?

2 Answers2