09 30 953029 21420119 86537117 618 25 00 32 33 46 36 3365 30 43 63 48 26 50 46 80 849193 59 10286153 88 5914 32 68 30 78 96 747 57 90 66 25 97 13 67 888623 08 80 52

number → string,
but looks good

Automatically detects and fixes floating-point precision issues • [GitHub]

THE PROBLEM

Floating-point arithmetic creates ugly precision artifacts that need smart formatting.

Native APIs require fixed parameters. For example `toFixed(4)` works for 123.45670001 but destroys 0.0000001 → “0.0000”.

Tricks like `+(n.toFixed(4)).toString()` remove extra 0s but still requires a fixed decimal length and it is still rounding (changing) the value.

`nstr()` automatically fixes potential number errors, and formats numbers to string without losing precisions.

0.1
.toString()
0.1
.toFixed(4)
0.1000
.toPrecision(4)
0.1000
nstr()
0.1
0.1 + 0.2
.toString()
0.30000000000000004
.toFixed(4)
0.3000
.toPrecision(4)
0.3000
nstr()
0.3
12.2 / 0.1
.toString()
121.99999999999999
.toFixed(4)
122.0000
.toPrecision(4)
122.0
nstr()
122
-0.0000001
.toString()
-1e-7
.toFixed(4)
-0.0000
.toPrecision(4)
-1.000e-7
nstr()
0
19.9 * 100
.toString()
1989.9999999999998
.toFixed(4)
1990.0000
.toPrecision(4)
1990
nstr()
1990
1.9999999999
.toString()
1.9999999999
.toFixed(4)
2.0000
.toPrecision(4)
2.000
nstr()
2
12345.60000002
.toString()
12345.60000002
.toFixed(4)
12345.6000
.toPrecision(4)
1.235e+4
nstr()
12345.6
9999999.123000001
.toString()
9999999.123000002
.toFixed(4)
9999999.1230
.toPrecision(4)
1.000e+7
nstr()
9999999.123
123.123456789
.toString()
123.123456789
.toFixed(4)
123.1235
.toPrecision(4)
123.1
nstr()
123.123456789

INTERACTIVE DEMO

Try these examples:

JavaScript Built-ins

.toString()
0.30000000000000004
.toFixed(3)
0.300
.toPrecision(5)
0.30000

nstr (Smart)

nstr()
0.3

USAGE

# Install
pnpm install nstr
# Basic usage
import nstr from 'nstr'
const result = nstr(0.1 + 0.2)
// result: "0.3"
# Advanced: configure max decimals, defaults to 10
nstr(Math.PI, { maxDecimals: 4 })
// result: "3.1416"
# Advanced: configure precision detection, defaults to 4
nstr(0.1239991, { threshold: 2 })
// result: "0.123" (detects shorter patterns)
nstr(0.1239991, { threshold: 5 })
// result: "0.1239991" (be more precise)

HOW DOES IT WORK?

Let's trace through the algorithm using `0.14499999582767487` as an example:

Step 1: Convert to fixed decimal
number.toFixed(maxDecimals=10) to avoid scientific notation and cap maximum digits
0.14499999582767487 → “0.1449999958”
Step 2: Detect consecutive patterns
Look for consecutive “0”s or “9”s longer than threshold=4
0.14499999
0.144999995 consecutive “9”s detected
58
Step 3: Truncate and clean up
Remove succeeding digits after the pattern
“0.1449999958” → “0.145”
Key parameters:
  • maxDecimals (default: 10) - Maximum decimal places to consider
  • threshold (default: 4) - Minimum consecutive 0s/9s to trigger cleanup

Built by [@shuding_] with [v0.app]