IOS-compute imprecise problems such as floatvalue,doublevalue data with high computational accuracy requirements

1. Problem Description: Double type 9 returned by server. 9,4. 94 of the number 29.89999961   ==> reserved 2 decimal point 9.  39.9000000000000003 ==> reserved 2 digits decimal point 9.  - 4 , other workaround: The data returned by the server is all set to type string, so the problem does not occur in memory, CPU calculation, float is more appropriate, in general, float

Double The difference between float is a double precision, a valid number 16 bits, and a float precision of 7 bits. But double consumes twice times the memory is float, double operation speed is much slower than float, C language math function name double and float different, do not write wrong, can use single precision do not use double precision (to save memory, speed up the operation speed).

One of the ways to resolve:

when designing to the calculation of money. But the use of double type data is unstable.1: Controls the precision of a double typeDoubleABC =0.1-0.01-0.00001009; Nsnumberformatter*NF =[[Nsnumberformatteralloc]init]; [NF Setmaximumintegerdigits:8]; NSNumber*number = [Nfnumberfromstring:[nsstring stringWithFormat:@"%.8LF", ABC]]; NSLog (@"Number is ---%lf", [Numberdoublevalue]);2: Calculation method for double type without loss of precision-(CGFloat) Addreebackmoneywithamount: (cgfloat) Amount Tomoney: (cgfloat) tomoney{nsstring*AMOUNTSTR = [NSString stringWithFormat:@"%.08LF", amount]; NSString*TOMONEYSTR = [NSString stringWithFormat:@"%.08LF", Tomoney]; Nsdecimalnumber*amountnum =[Nsdecimalnumber Decimalnumberwithstring:amountstr]; Nsdecimalnumber*tomoneynum =[Nsdecimalnumber Decimalnumberwithstring:tomoneystr]; DoubleXiaofee =0.001210000; Nsdecimalnumber*feenum = [Nsdecimalnumber decimalnumberwithstring:[nsstring stringWithFormat:@"%.8LF", Xiaofee]]; Nsdecimalnumber*resultnum =[Amountnum Decimalnumberbysubtracting:tomoneynum]; Nsdecimalnumber*subtracfeenum =[Resultnum Decimalnumberbysubtracting:feenum]; return[Subtracfeenum doublevalue];;}

Give me a chestnut: calculate 0.1*999999 and see what happens?

-(void) testdecimalnumber {     double0.01;      Double 999999 ;      double d3 = d1 * D2;       NSLog (@ "%@", N3);  }

And we think of the results: 9999.99 different

The solution to this problem is as follows:

-(void) testdecimalnumber {     nsdecimalnumber* n1 = [Nsdecimalnumber       Decimalnumberwithstring:[nsstring    stringWithFormat:@ "%f", D1]];    Nsdecimalnumber* n2 = [Nsdecimalnumber decimalnumberwithstring:[nsstring stringwithformat:@ "%f  ", D2]];    Nsdecimalnumber* n3 = [N1 DECIMALNUMBERBYMULTIPLYINGBY:N2]; NSLog (@ "%@", number);}

You can convert a float, double to an object of type Nsdecimalnumber in the calculation. You can perform operations of + 、-、 *,/.

Usage Rounding:

-(nsstring*) notrounding: (float) Price afterpoint: (nsinteger) position{    Nsdecimalnumberhandler * Roundinghandler = [Nsdecimalnumberhandler decimalnumberhandlerwithroundingmode:nsroundplain scale: Position Raiseonexactness:no Raiseonoverflow:no Raiseonunderflow:no Raiseondividebyzero:no];    Nsdecimalnumber* returnnumber;    Nsdecimalnumber* ouncesdecimal = [[Nsdecimalnumber alloc]initwithfloat:price];     = [Ouncesdecimal decimalnumberbyroundingaccordingtobehavior:roundinghandler];     return [NSString stringWithFormat:@ "%@", Returnnumber];}

Test data: 1.235

Results:

IOS-compute imprecise problems such as floatvalue,doublevalue data with high computational accuracy requirements