{{ summaryHeading }}
{{ primaryDisplay }}
{{ summaryLine }}
{{ countBadge }} {{ notationBadge }} {{ ambiguityBadge }} {{ roundingBadge }}
Use decimal or scientific notation, such as 0.004500, 750., or 1.230e-4.
Choose the significant-figure target for the preview and error curve.
sig figs
Conventional mode treats plain trailing integer zeros as placeholders; measured mode counts them.
This changes display format only; the count and rounding target stay the same.
Use 4-15 rows; this affects only the exportable rounding plan.
rows
Check Value Readout Copy
{{ row.check }} {{ row.value }} {{ row.readout }}
Token Place Role Counted Rule Copy
{{ row.token }} {{ row.place }} {{ row.role }} {{ row.counted }} {{ row.rule }}
Sig figs Rounded value Absolute error Relative error Readout Copy
{{ row.sigFigs }} {{ row.roundedValue }} {{ row.absoluteError }} {{ row.relativeError }} {{ row.readout }}
The rounding error curve needs a nonzero numeric value that fits browser number range.

          

        

      

    

  

                

                

                    

                        

                            Tags: Education
                        

                        

                            Export to PDF
                            Fullscreen
                        

                    

                

            

        

    


                    
        

        

        

            

            
        

        


                                                                                                                                                                                                                                                        
        

        

            
            

                Customize

                

                    

                        

                            
                            
                        


                        

                            
                            
                        


                        

                            
                            
                        

                    


                    

                        Advanced

                        

                            

                                
                                

                                    
                                    
                                

                            


                            

                                
                                

                                    
                                    
                                

                            


                            

                                
                                

                                    
                                    :
                                    
                                

                            


                            

                                
                                

                                    
                                    
                                

                            


                            

                                

                                    
                                    
                                

                            


                            

                                

                                    
                                    
                                

                            


                            

                                
                                
                            


                            

                                
                                
                            

                        

                    

                

            

            
        


        

            
            
            
        


        
        

    

    

        

  

    
    
Significant figures are the written digits that carry measurement precision. Nonzero digits count, zeros between counted digits count, and leading zeros usually mark decimal place rather than precision. Trailing zeros need more care because the same numeric value can be written to show different levels of measurement detail.


    
The count matters when a value will be rounded, compared with a lab note, copied into a calculation, or reported with a measurement. Writing 0.004500 says something different from writing 0.0045, even though both have the same numeric value. The extra zeros show that the value was kept to the hundred-thousandths and millionths places.


    

      Digit role diagram for 0.004500 showing leading zeros, counted nonzero digits, counted decimal zeros, and equivalent scientific notation.
    


    
Scientific notation is often the clearest way to remove ambiguity. The coefficient shows the significant figures, and the exponent moves the decimal point. For example, 1.2e+3, 1.20e+3, and 1.200e+3 all represent 1200 numerically, but they report two, three, and four significant figures.


    
A significant-figure count does not prove that a measurement was made carefully. It only reads the precision implied by the written number. If the value came from a sensor, answer key, spreadsheet, or rounded report, the source convention still matters.

  






  

    
Technical Details:

    
Significant-figure counting works on the written coefficient, not on the numeric value alone. A parser that turns 0.004500 into the number 0.0045 has already lost the two trailing zeros that express precision. The written text must be preserved long enough to classify each digit.


    
The first nonzero digit anchors the count. Digits before it are leading zeros and only locate the decimal point. After the anchor, nonzero digits count, zeros between counted digits count, and trailing zeros count when the written form states that they are measured. A decimal point, scientific notation coefficient, or explicit measured-zero convention can make trailing zeros count.


    
Rule Core:

    
The count can be described as a span from the first nonzero digit to the last counted digit.


    

      
        
          
            c
            =
            0 if no nonzero digit appears
          
          
            c
            =
            j-i+1 otherwise
          
          
            e
            =
            |r-x||x|×100%
          
        
      
    


    
Here i is the first nonzero digit index, j is the last digit counted under the selected zero rule, c is the significant-figure count, x is the parsed numeric value, r is the rounded preview, and e is the relative rounding error percent when the original numeric value is nonzero.


    

      

        
        
          

            Digit role
            Counts?
            Rule used
            Example
          

        
        
          

            Leading zero
            No
            Zeros before the first nonzero digit locate decimal place.
            0.004500 has three leading zeros before 4.
          

          

            Nonzero digit
            Yes
            Every nonzero digit in the coefficient is significant.
            4 and 5 count in 0.004500.
          

          

            Captive zero
            Yes
            A zero between counted digits is part of the measured value.
            70.607 keeps both interior zeros.
          

          

            Decimal trailing zero
            Yes
            A zero after the decimal point and after a nonzero digit preserves written precision.
            55.0 and 1.230e-4 include the final zero.
          

          

            Plain integer trailing zero
            Depends
            Without a decimal point, notation, or convention, a trailing integer zero may be only a placeholder.
            1200 is treated as two significant figures in conventional mode.
          

          

            Scientific exponent
            No
            The exponent scales the value and does not add significant figures.
            1.230e-4 has four significant figures, not five.
          

        
      

    


    
Rounding to a target count uses the significant digits after the first nonzero digit. Digits beyond the target are dropped, and the retained digit is increased when the next dropped digit is 5 or greater. If a carry changes the leading digit, the power of ten shifts by one place. The preview can display a short decimal form or scientific notation, but the scientific equivalent remains available for clarity.


    

      

        
        
          

            Item
            Accepted rule
            Why it matters
          

        
        
          

            Number text
            Plain decimal or scientific notation such as 0.004500, 750., 1.230e-4, or 1.230x10^-4.
            The written decimal point, coefficient, and trailing zeros decide the count.
          

          

            Text cleanup
            Spaces, commas, and underscores are ignored before parsing.
            Copied values such as 1,200 can be read while the written digits remain available.
          

          

            Round to
            Whole-number target from 1 to 15 significant figures.
            Controls the rounded preview and the highlighted target in the error curve.
          

          

            Integer trailing zeros
            Conventional placeholders or Count as measured zeros.
            Changes how plain integers such as 1200 are counted.
          

          

            Rounded format
            Auto, preserve clarity, Scientific notation, or Decimal if short.
            Changes the display form without changing the count or numeric rounded value.
          

          

            Plan depth
            Whole-number row count from 4 to 15.
            Controls how many target counts appear in the rounding plan and error curve.
          

        
      

    


    
Zero-only values are treated separately because a written 0 does not reveal measurement precision by itself. The count is reported as zero significant figures, and the summary warns that measurement context is needed. Very large or tiny exponents can still be counted from text even when a browser numeric value cannot support an error curve.

  






  

    
Everyday Use & Decision Guide:

    
Start by typing the number exactly as it was reported. Keep the decimal point, trailing zeros, and scientific notation because those marks are part of the precision statement. A value copied as 0.004500 should not be shortened to 0.0045 before checking it.


    
Leave Integer trailing zeros on Conventional placeholders for most textbook and report-review work. Switch to Count as measured zeros only when a measurement convention says that a plain integer such as 1200 was recorded to the ones place. If the source can be rewritten, 1200. or 1.200e+3 communicates the same intent more clearly.


    
    
      
  • Use Figure Audit to confirm the count, scientific notation, first significant place, last significant place, rounded preview, and integer-zero convention.
    • 
      
  • Use Digit Ledger when a single zero is in question. Each digit shows its place, role, counted status, and rule.
    • 
      
  • Use Rounding Plan to compare several target counts before choosing how many digits to report.
    • 
      
  • Use Rounding Error Curve when a nonzero numeric value fits the browser number range and you want to see how relative error falls as more digits are kept.
    • 
      
  • Use JSON when you need a structured record of the input, count, parsed coefficient, rounded preview, ledger, and errors.
    • 
    


    
Do not treat the rounded preview as a full significant-figures rule for every calculation. Addition, subtraction, multiplication, and division have their own reporting rules. This page is strongest when the immediate question is how many significant figures are visible in one written number and how that same number would look after rounding to a chosen count.

  






  

    
Step-by-Step Guide:

    
Work from the written value to the count, then use the ledger and rounding views to check any questionable zeros.


    
    
      
  1. Enter the exact reported value in Number text. Accepted forms include 0.004500, 750., 1.230e-4, and 1.230x10^-4.
    2. 
      
  2. Set Round to from 1 to 15. The summary badge updates to show the selected target, and Figure Audit shows the rounded preview.
    3. 
      
  3. Open Advanced when a plain integer ends in zeros. Keep Conventional placeholders for ordinary 1200, or choose Count as measured zeros when those zeros were reported as measured digits.
    4. 
      
  4. Choose Rounded format if the preview needs a specific form. Auto, preserve clarity uses scientific notation when decimal output would hide the target count.
    5. 
      
  5. Read the summary first. It should show a count such as 4 sig figs, a notation badge such as Decimal shown or Scientific input, and an ambiguity badge such as Rules resolved or Trailing zeros ambiguous.
    6. 
      
  6. Open Digit Ledger if the count surprises you. A zero marked Leading zero or Placeholder zero is not counted; a zero marked Significant zero is counted.
    7. 
      
  7. Use Rounding Plan and Rounding Error Curve to compare target counts. If the curve panel says it needs a nonzero numeric value that fits browser number range, use the table and scientific equivalent instead.
    8. 
      
  8. If a red validation alert appears, fix the number format first. The usual correction is to enter one plain number or scientific-notation value, not surrounding words or multiple numbers.
    9. 
    

  






  

    
Interpreting Results:

    
The headline count is the number of significant figures visible under the selected convention. The best verification cue is the Digit Ledger, because it explains why each digit was counted or ignored. If the count changes only when Integer trailing zeros changes, the written value is ambiguous rather than mathematically different.


    

      

        
        
          

            Output
            What it means
            Check before trusting it
          

        
        
          

            Significant figures
            Count of written digits from the first nonzero digit through the last counted digit.
            Confirm the decimal point, exponent, and integer-zero convention match the source value.
          

          

            Scientific notation
            Coefficient and power-of-ten form that preserves the counted digits.
            Use it to make a plain integer's precision explicit when trailing zeros are unclear.
          

          

            First significant place
            Place value of the first counted digit, such as thousandths or hundreds.
            Make sure leading zeros are being used only to locate the decimal point.
          

          

            Last significant place
            Place value of the last counted digit, which is the implied reporting precision.
            Review trailing zeros carefully because this place can change with the selected convention.
          

          

            Relative error
            Percent difference between the rounded preview and the parsed numeric value.
            It is not available for zero values and can be unavailable when the numeric value exceeds browser range.
          

        
      

    


    
The main false-confidence trap is reading a clean count as proof of real measurement certainty. A spreadsheet value, rounded estimate, exact count, and lab measurement may all look similar as text. Use the count to audit the written number, then compare it with the original measurement convention before reporting the result.

  






  

    
Worked Examples:


    
Decimal Zeros That Count:

    
With Number text set to 0.004500 and Round to set to 3, the summary shows 4 sig figs. Figure Audit reports Scientific notation as 4.500e-3, First significant place as thousandths, and Last significant place as millionths. The rounded preview is 0.00450, so the two final zeros in the original value are not ignored.


    
A Plain Integer With Ambiguous Zeros:

    
Enter 1200 with Integer trailing zeros left on Conventional placeholders. The summary shows 2 sig figs and the ambiguity badge says Trailing zeros ambiguous. Digit Ledger marks the two zeros as Placeholder zero. Switch to Count as measured zeros, and the same text counts as 4 sig figs with scientific notation 1.200e+3.


    
Scientific Notation With a Scale Exponent:

    
For 1.230e-4 and Round to 2, the count is 4 sig figs. The exponent row in Digit Ledger is marked Exponent and Counted is No. The rounded preview is 1.2e-4, with a relative error of about 2.439% in Rounding Plan.


    
Fixing a Format Error:

    
If Number text contains words or more than one value, the result area is replaced by a red validation alert. A phrase such as about 0.004500 m should be shortened to 0.004500. After the alert clears, return to Figure Audit and confirm that the input token still includes the decimal places you meant to preserve.

  






  

    
FAQ:


    

      Why does 0.004500 have four significant figures?
      
The leading zeros before 4 locate the decimal point and are not counted. The 4, 5, and the two trailing zeros after the decimal point are counted, so Digit Ledger marks four digits as significant.

    


    

      Why can 1200 count as two or four significant figures?
      
Plain trailing integer zeros are ambiguous. In Conventional placeholders mode, 1200 counts as two significant figures. In Count as measured zeros mode, the two zeros count too, so the result becomes four significant figures.

    


    

      Does the exponent count in scientific notation?
      
No. The exponent changes scale only. For 1.230e-4, the coefficient 1.230 supplies the four significant figures, and the exponent row is shown as not counted.

    


    

      Why did the rounded preview switch to scientific notation?
      
Auto, preserve clarity uses scientific notation when a decimal display would be very long or could hide the intended significant-figure target. Choose Scientific notation when you always want that form.

    


    

      Why is the rounding error curve missing?
      
The curve needs a nonzero numeric value that fits browser number range. Zero-only values, invalid inputs, and extreme exponents can still produce count information, but the relative error curve may not have usable numeric points.

    


    

      Are my numbers sent to a server for counting?
      
The count, digit ledger, rounding plan, and JSON record are produced in the page. No server-side helper is used for the calculation, though the chart dependency can load from a public content delivery network before the error curve is drawn.

    

  






  

    
Glossary:

    

      
Significant figure

      
A written digit that communicates measurement precision.

      
Leading zero

      
A zero before the first nonzero digit; it locates decimal place and is not counted.

      
Trailing zero

      
A zero after the last nonzero digit; it may count when the notation or convention shows measured precision.

      
Captive zero

      
A zero between nonzero significant digits, counted as part of the measured value.

      
Scientific notation

      
A coefficient and power-of-ten form where the coefficient carries the significant figures and the exponent carries scale.

      
Relative error

      
The absolute rounding difference divided by the original numeric value, shown as a percent when available.

    

  






  

    
References: