Cannot explicitly convert string to float

Question

I am using the following code to format currency in my unity game. But there is some error messages saying "Cannot explicitly convert string to float"

    public float FormatAmount(float amount)
    {
        if (amount <= 999) {
            return amount;
        }
        // Thousands 106
        else if (amount >= 1000 && amount <= 999999) {
            return (amount / 1000 + "K");
        }
        // Millions 106
        else if (amount >= 1000000 && amount <= 999999999) {
            return (amount / 1000000 + "M");
        }

        // Billions 109
        else if (amount >= 1000000000 && amount <= 999999999999) {
            return (amount / 1000000000 + "B");
        }

        // Trillions 1012
        else if (amount >= 1000000000000 && amount <= 999999999999999) {
            return (amount / 1000000000000 + "T");
        }

        // Quadrillions 1015
        else if (amount >= 1000000000000000 && amount <= 999999999999999999) {
            return (amount / 1000000000000000 + "Quad");
        }
        // Quintillions 1018
        else if (amount >= 1000000000000000000 && amount <= 999999999999999999) {
            return (amount / 1000000000000000000 + "Quin");
        } else {
            return amount;
        }
    }

You are returning a string, but your method signature is to return a float. Stop appending a string to the return values. — maccettura, Jun 18 '18 at 15:57
`(amount / 1000 + "K");`, `(amount / 1000000 + "M")` and so on. You can't do that. You are trying to add string to float. What do you expect to happen? — Programmer, Jun 18 '18 at 15:58
Question, can you not use the culture variance for determining currency? — Greg, Jun 18 '18 at 16:02
@Programmer: you can do that - the numeric part is automatically converted to a string - if amount is 1234 the result is "1.234K" — PaulF, Jun 18 '18 at 16:05
@PaulF I thought OP wanted to add the string value to the float value — Programmer, Jun 18 '18 at 16:06
@Programmer: In that case it would be wrong, but I read it as adjusting the value & appending a string suffix for a display output string. — PaulF, Jun 18 '18 at 16:08
I also spotted the final comparison is 1 9 short - should be _"amount <= 9999999999999999999"_. Also with values that big, float may not be suitable : _"FormatAmount(1234567891234567894)"_ results in "1.234568Quin" for float, "1.23456789123457Quin" for float & "1.234567891234567894Quin" for decimal - so if full accuracy is required the parameter should be passed as decimal. — PaulF, Jun 18 '18 at 16:18
Also, if you are making an idle game and want stupid huge values, I suggest getting a `BigInteger` class like [this one](https://numerics.codeplex.com/). I use [this function](https://github.com/Draco18s/IdleArtificer/blob/master/Assets/draco18s/artificer/game/Main.cs#L839) to trim down to a maximum number of displayable digits. — Draco18s no longer trusts SE, Jun 18 '18 at 19:36

AlexT · Accepted Answer · 2018-06-19T12:48:36.380

2

There are a couple of problems with the code that you have provided. The method signature is declared as having a float output:

public float FormatAmount(float amount)

The float is the output and is defined as a float. And the second float is the input and it required a float (this one is ok if you call the function while providing a float)

But, in your returns, you have

return amount - which will pass a float.
return (amount / 1000 + "K"); - which will pass a string.

There are 2 possibilities here.

If you need a float as the output of the function, you need to only return a float. That means that you will not have the +"letter".
If you want to keep the letters and thus return a string, you need to make sure that you are only passing a string. That means that you can keep return (amount / 1000 + "K");, but you will also need to add .toString() to the return where you have only a float. So in the end it will look like this:

public string FormatAmount(float amount)
    {
        if (amount <= 999) {
            return amount.toString();
        }
        else if (amount >= 1000 && amount <= 999999) {
            return (amount / 1000 + "K");
        }
    }

You might also want to take PaulF's comment into account regarding the limitations of float, and Draco18's solution of using an external numeric library for values that large.

edited Jun 19 '18 at 12:48

answered Jun 18 '18 at 15:58

AlexT

2,524
1
11
23

This is not the issue. – Programmer Jun 18 '18 at 15:59
Care to explain how is that not the issue? He declared a float as the return param, and the tries to return a string. Either he gets ring of the +"String" or changes the return type. Or am I wrong ? – AlexT Jun 18 '18 at 16:02
If OP does require a string returning - that does actually solve the problem other than the two "return amount;" needs to change to "return amount.ToString();" – PaulF Jun 18 '18 at 16:02
PaulF made a comment I wanted to make. It depends on what wants to returning and simply doing what's in your answer wouldn't solve all the problem. – Programmer Jun 18 '18 at 16:04
Considering that OP is saying that the purpose is to "format currency". It is ok (I believe) to assume that he does actually want a string. edit: As to "wouldn't solve all the problem" you are assuming the rest of his code, which is the same thing I am doing, only in the opposite direction. – AlexT Jun 18 '18 at 16:09
The first condition would also need to return `amount.ToString();`. Otherwise, this sounds like what the OP was looking for to me. – Rufus L Jun 18 '18 at 16:13
@RufusL: the last condition as well - I already made that comment. – PaulF Jun 18 '18 at 16:19
See my comment to the question regarding accuracy of the return string - if full accuracy is expected the parameter needs to be decimal. – PaulF Jun 18 '18 at 16:20

score 0 · Answer 2 · answered Jun 18 '18 at 16:00

0

It's correct, you are concatenate a string and your function returns float. Change the return method type.

answered Jun 18 '18 at 16:00

Leandro Sousa

42
3

Cannot explicitly convert string to float

2 Answers2