指数精度[打印双精度，小数]

Question

有没有一种打印精度达到指数级的方法？我需要用微小的数字填充一些数据，然后将它们进行比较……

我有这样的想法：

(0) - Mesured estimatedTime for multithread in block: 1.4917951E16 s
(1) - Mesured estimatedTime for multithread in block: 7.531891E15 s
(2) - Mesured estimatedTime for multithread in block: 2.9287E13 s
(3) - Mesured estimatedTime for multithread in block: 3.28478435E17 s
(4) - Mesured estimatedTime for multithread in block: 6.038E12 s

我想这样打印：

(0) - Mesured estimatedTime for original in block: 0000.15595175E15 s           
(1) - Mesured estimatedTime for original in block: 0007.335638E15 s             
(2) - Mesured estimatedTime for original in block: 0416.66E15 s                 
(3) - Mesured estimatedTime for original in block: 0000.0390156852E15 s         
(4) - Mesured estimatedTime for original in block: 6642.0E15 s

我知道你可以像这样强迫它：

// Force minimum number of digits to left and right of decimal point
formatter = new DecimalFormat("0.0E0");
s = formatter.format(-1234.567);         // -1.2E3

但我不知道如何才能将它强制到Expo-Precision：( help!！:/

Answer 1

看起来你想暗示1e15的精度。

for (double d : new double[]{1.4917951E16, 7.531891E15, 2.9287E13, 3.28478435E17, 6.038E12})
    System.out.printf("%11fE15%n", d/1e15);

打印

  14.917951E15
   7.531891E15
   0.029287E15
 328.478435E15
   0.006038E15

如果范围大得多，可以使用DecimalFormat

DecimalFormat df = new DecimalFormat();
df.setMaximumFractionDigits(300);
df.setGroupingUsed(false);
for(double d = 1 ; d <= 1e30; d *= 1e5) {
    System.out.println(df.format(d/1e15)+"E15");
}

打印

0.000000000000001E15
0.0000000001E15
0.00001E15
1E15
100000E15
10000000000E15