Are you sure?
This action might not be possible to undo. Are you sure you want to continue?
B U R E A U O F I N D I A N S T A N D A R D S
MANAK BHAVAN, 9 BAHADUR SHAH ZAFAR MARG
NEW DELHI 110002
I S : 2  1960
(Reaffi r med 1990)
Edi t i on 2.3
(200008)
Pr i ce Gr oup 3
I ndi an St andar d
RULES FOR ROUNDI NG OFF
NUMERI CAL VALUES
( Revi sed )
(I ncor por at i ng Amendment Nos. 1, 2 & 3)
UDC 511.135.6
( Reaffirmed 2000 )
I S : 2  1960
B U R E A U O F I N D I A N S T A N D A R D S
MANAK BHAVAN, 9 BAHADUR SHAH ZAFAR MARG
NEW DELHI 110002
I ndi an St andar d
RULES FOR ROUNDI NG OFF
NUMERI CAL VALUES
( Revi sed )
Engi neer i ng St andar ds Sect i onal Commi t t ee, EDC 1
Chai r man
DR K. S. KRI SHNAN Counci l of Sci ent i fi c & I ndust r i al Resear ch,
New Del hi
Member s
SHRI PREM PRAKASH ( Al ter nat e t o
Dr K. S. Kr i shnan )
ASSI STANT DI RECTOR Resear ch, Desi gn & St andar di zat i on Or gani zat i on
(Mi ni st r y of Rai l ways)
SHRI BALESHWAR NATH Cent r al Boar d of I r r i gat i on & Power , New Del hi
DEPUTY DI RECTOR GENERAL OF
OBSERVATORI ES
Di r ect or at e Gener al of Obser vat or i es (Mi ni st r y of
Tr anspor t & Communi cat i ons), New Del hi
DI RECTOR Engi neer i ng Resear ch Depar t ment , Gover nment
of Andhr a Pr adesh
SHRI S. B. JOSHI I nst i t ut i on of Engi neer s (I ndi a), Cal cut t a
SHRI R. N. KAPUR I ndi an Engi neer i ng Associ at i on, Cal cut t a
DR R. S. KRI SHNAN I ndi an I nst i t ut e of Sci ence, Bangal ore
SHRI S. R. MEHRA Counci l of Sci ent i fi c & I ndust r i al Resear ch, New
Del hi
SHRI S. N. MUKERJI
SHRI K. D. BHATTACHARJEE
( Al ter nat e )
Gover nment Test House, Cal cut t a
DR B. R. NI JHAWAN Counci l of Sci ent i fi c & I ndust r i al Resear ch, New
Del hi
SHRI V. R. RAGHAVAN Cent r al Wat er & Power Commi ssi on, New Del hi
BRI G J. R. SAMSON
LTCOL R. JANARDHANAM
( Al ter nat e )
Cont r ol l er of Devel opment (Ar mament s) (Mi ni st r y
of Defence)
( Cont i nued on page 2 )
I S : 2  1960
2
( Cont i nued fr om page 1 )
Member s
SHRI R. N. SARMA Di r ect or at e Gener al of Suppl i es & Di sposal s
(Mi ni st r y of Works, Housi ng & Suppl y)
SHRI J. M. SI NHA Engi neer i ng Associ at i on of I ndi a, Cal cut t a
SHRI J. M. TREHAN
SHRI T. N. BHARGAVA
( Al ter nat e )
Mi ni st r y of Tr anspor t & Communi cat i ons (Roads
Wi ng)
LTGEN H. WI LLI AMS Counci l of Sci ent i fi c & I ndust r i al Resear ch,
New Del hi
DR LAL C. VERMAN ( Exoffi ci o )
SHRI J. P. MEHROTRA
( Al ter nat e )
Di r ect or , I SI
Deput y Di r ect or (Engg), I SI
Secr et ar i es
DR A. K. GUPTA Assi st ant Di r ect or (St at ), I SI
SHRI B. N. SI NGH Ext r a Assi st ant Di rect or (St at ), I SI
I S : 2  1960
3
I ndi an St andar d
RULES FOR ROUNDI NG OFF
NUMERI CAL VALUES
( Revi sed )
0. F O R E W O R D
0.1 Thi s I ndi an St andar d (Revi sed) was adopt ed by t he I ndi an
St andar ds I nst i t ut i on on 27 Jul y 1960, aft er t he dr aft fi nal i zed by t he
Engi neer i ng St andar ds Sect i onal Commi t t ee had been appr oved by t he
Engi neer i ng Di vi si on Counci l .
0.2 To r ound off a val ue i s t o r et ai n a cer t ai n number of fi gur es, counted
fr om the l eft , and dr op t he ot her s so as t o gi ve a mor e r at i onal for m t o t he
val ue. As t he r esul t of a t est or of a cal cul at i on i s gener al l y r ounded off
for t he pur pose of r epor t i ng or for dr aft i ng speci fi cat i ons, i t i s necessar y
t o pr escr i be r ul es for ‘r oundi ng off’ numer i cal val ues as al so for deci di ng
on ‘t he number of fi gur es’ t o be r et ai ned.
0.3 Thi s st andar d was or i gi nal l y i ssued i n 1949 wi t h a vi ew t o pr omot i ng
t he adopt i on of a uni for m pr ocedur e i n r oundi ng off numer i cal val ues.
However , t he r ul es gi ven r efer r ed onl y t o uni t fi neness of r oundi ng
( see 2.3 ) and i n cour se of year s t he need was fel t t o pr escr i be r ul es for
r oundi ng off numer i cal val ues t o fi neness of r oundi ng ot her t han uni t y.
Mor eover , i t was al so fel t t hat t he di scussi on on t he number of fi gur es t o
be r et ai ned as gi ven i n t he ear l i er ver si on r equi r ed fur t her el uci dat i on.
The pr esent r evi si on i s expect ed t o ful fi l t hese needs.
0.4 I n pr epar i ng t hi s st andar d, r efer ence has been made t o t he
fol l owi ng:
I S : 7871956 GUI DE FOR I NTERCONVERSI ON OF VALUES FROM ONE
SYSTEM OF UNI TS TO ANOTHER. I ndi an St andar ds I nst i t ut i on.
I S : 1890 (PART 0)1995/I SO 310 : 1992 QUANTI TI ES AND UNI TS :
PART 0 GENERAL PRI NCI PLES ( FI RST REVI SI ON ).
BS 1957 : 1953 PRESENTATI ON OF NUMERI CAL VALUES (FI NENESS OF
EXPRESSI ON; ROUNDI NG OF NUMBERS). Br i t i sh St andar ds
I nst i t ut i on.
AMERI CAN STANDARD Z 25.11940 RULES FOR ROUNDI NG OFF
NUMERI CAL VALUES. Amer i can St andar ds Associ at i on.
ASTM DESI GNATI ON : E 2950 RECOMMENDED PRACTI CE FOR
DESI GNATI NG SI GNI FI CANT PLACES I N SPECI FI ED VALUES. Amer i can
Soci et y for Test i ng and Mat er i al s.
JAMES W. SCARBOROUGH. Numer i cal Mat hemat i cal Anal ysi s
Bal t i mor e. The John Hopki ns Pr ess, 1955.
I S : 2  1960
4
0.5 Thi s edi t i on 2.3 i ncor por at es Amendment No. 1 (Febur ar y 1997),
Amendment No. 2 (Oct ober 1997) and Amendment No. 3 (August 2000).
Si de bar i ndi cat es modi fi cat i on of t he t ext as t he r esul t of
i ncor por at i on of t he amendment s.
1. SCOPE
1.1 Thi s st andar d pr escr i bes r ul es for r oundi ng off numer i cal val ues for
t he pur pose of r epor t i ng r esul t s of a t est , an anal ysi s, a measur ement or
a cal cul at i on, and t hus assi st i ng i n dr aft i ng speci fi cat i ons. I t al so makes
r ecommendat i ons as t o t he number of fi gur es t hat shoul d be r et ai ned i n
cour se of comput at i on.
2. TERMI NOLOGY
2.0 For t he pur pose of t hi s st andar d, t he fol l owi ng defi ni t i ons shal l
appl y.
2.1 Number of Deci mal Places — A val ue i s sai d t o have as many
deci mal pl aces as t her e ar e number of fi gur es i n t he val ue, count i ng
fr om t he fi r st fi gur e aft er t he deci mal poi nt and endi ng wi t h t he l ast
fi gur e on t he r i ght .
NOTE 1 — For t he pur pose of t hi s st andard, t he expr essi on 10.32 × 10
3
shoul d
be t aken t o consi st of t wo par t s, t he val ue pr oper whi ch i s 10.32 and t he uni t of
expressi on for t he val ue, 10
3
.
2.2 Number of Si gni fi cant Fi gur es — A val ue i s sai d t o have as many
si gni fi cant fi gur es as t her e ar e number of si gni fi cant di gi ts ( see Not e 2 )
i n t he val ue, count i ng fr om t he l eft most nonzer o di gi t and endi ng wi t h
t he r i ght most di gi t i n t he val ue.
Exampl es:
Val ue Deci mal Pl aces
0.029 50 5
21.029 5 4
2 000.000 001 6
291.00 2
10.32 × 10
3
2
( see Not e 1 )
Exampl es:
Val ue Si gni fi cant Fi gur es
0.029 500 5
0.029 5 3
10.029 5 6
2 000.000 001 10
5 677.0 5
567 700 6
56.77 × 10
2
4
0 056.770 5
3 900 4
( see Not e 3 )
I S : 2  1960
5
NOTE 2 — Any of t he di gi t s, 1, 2, 3,............9 occur r i ng i n a val ue shal l be a
si gni fi cant di gi t (s); and zer o shal l be a si gni fi cant di gi t onl y when i t i s pr eceded by
some ot her di gi t (except i ng zer os) on i t s l eft . When appear i ng i n t he power of 10 t o
i ndi cat e t he magni t ude of t he uni t i n t he expr essi on of a val ue, zer o shal l not be a
si gni f i cant di gi t .
NOTE 3 — Wi t h a vi ew t o r emovi ng any ambi gui t y regardi ng t he si gni fi cance of
t he zer os at t he end i n a val ue l i ke 3 900, i t woul d be al ways desi r abl e t o wri t e t he
val ue i n t he power oft en not at i on. For exampl e, 3 900 may be wr i t t en as 3.9 × 10
3
,
3.90 × 10
3
or 3.900 × 10
3
dependi ng upon t he l ast fi gur e(s) i n t he val ue t o whi ch i t
i s desi r ed t o i mpar t si gni fi cance.
2.3 Fi neness of Roundi ng — The uni t t o whi ch a val ue i s r ounded off.
For exampl e, a val ue may be r ounded t o t he near est 0.000 01, 0.000 2,
0.000 5, 0.001, 0.002 5, 0.005, 0.01, 0.07, 1, 2.5, 10, 20, 50, 100 or any
ot her uni t dependi ng on t he fi neness desi r ed.
3. RULES FOR ROUNDI NG
3.0 The r ul e usual l y fol l owed i n r oundi ng off a val ue t o uni t fi neness of
r oundi ng i s t o keep unchanged t he l ast fi gur e r et ai ned when t he fi gur e
next beyond i s l ess t han 5 and t o i ncr ease by 1 t he l ast fi gur e r et ai ned
when t he fi gur e next beyond i s mor e t han 5. Ther e i s di ver si t y of
pr act i ce when t he fi gur e next beyond t he l ast fi gur e r et ai ned i s 5. I n
such cases, some comput er s ‘r ound up’, t hat i s, i ncr ease by 1, t he l ast
fi gur e r et ai ned; ot her s ‘r ound down’, t hat i s, di scar d ever yt hi ng beyond
t he l ast fi gur e r et ai ned. Obvi ousl y, i f t he r et ai ned val ue i s al ways
‘r ounded up’ or al ways ‘r ounded down’, t he sum and t he aver age of a
ser i es of val ues so r ounded wi l l be l ar ger or smal l er t han t he
cor r espondi ng sum or aver age of t he unr ounded val ues. However , i f
r oundi ng off i s car r i ed out i n accor dance wi t h t he r ul es st at ed i n 3.1 i n
one st ep ( see 3.3 ), t he sum and t he aver age of t he r ounded val ues woul d
be mor e near l y cor r ect t han i n t he pr evi ous cases ( see Appendi x A ).
3.1 Rounding Off t o Uni t Fi neness — I n case t he fi neness of
r oundi ng i s uni t y i n t he l ast pl ace r et ai ned, t he fol l owi ng r ul es (except
i n 3.4) shal l be fol l owed:
Rul e I — When t he fi gur e next beyond t he l ast fi gur e or pl ace t o be
r et ai ned i s l ess t han 5, t he fi gur e i n t he l ast pl ace r et ai ned shal l be l eft
unchanged.
Rul e I I — When t he fi gur e next beyond t he l ast fi gur e or pl ace t o be
r et ai ned i s mor e t han 5 or i s 5 fol l owed by any fi gur es ot her t han zer os,
t he fi gur e i n t he l ast pl ace r et ai ned shal l be i ncr eased by 1.
Rul e I I I — When t he fi gur e next beyond t he l ast fi gur e or pl ace t o be
r et ai ned i s 5 al one or 5 fol l owed by zer os onl y, t he fi gur e i n t he l ast pl ace
r et ai ned shal l be (a) i ncr eased by 1 i f i t i s odd and (b) l eft unchanged i f
even (zer o woul d be r egar ded as an even number for t hi s pur pose).
I S : 2  1960
6
Some exampl es i l l ust r at i ng t he appl i cat i on of Rul es I t o I I I ar e gi ven i n
Tabl e I .
3.1.1 The r ul es for r oundi ng l ai d down i n 3.1 may be ext ended t o appl y
when t he fi neness of r oundi ng i s 0.10, 10, 100, 1 000, et c. For
exampl e, 2.43 when r ounded t o fi neness 0.10 becomes 2.40.
Si mi l ar l y, 712 and 715 when r ounded t o t he fi neness 10 become 710
and 720 r espect i vel y.
3.2 Roundi ng Off t o Fineness Ot her than Unit y — I n case t he
fi neness of r oundi ng i s not uni t y, but , say, i t i s n, t he gi ven val ue shal l
be r ounded off accor di ng t o t he fol l owi ng r ul e:
Rul e I V — When r oundi ng t o a fi neness n, ot her t han uni t y, t he gi ven
val ue shal l be di vi ded by n. The quot i ent shal l be r ounded off t o t he
near est whol e number i n accor dance wi t h t he r ul es l ai d down i n 3.1 for
uni t fi neness of r oundi ng. The number so obt ai ned, t hat i s t he r ounded
quot i ent , shal l t hen be mul t i pl i ed by n t o get t he fi nal r ounded val ue.
Some exampl es i l l ust r at i ng t he appl i cat i on of Rul e I V ar e gi ven i n
Tabl e I I .
NOTE 4 — The r ul es for r oundi ng off a val ue t o any fi neness of r oundi ng, n, may
al so be st at ed i n l i ne wi t h t hose for uni t fi neness of r oundi ng ( see 3.1 ) as fol l ows:
Di vi de t he gi ven val ue by n so t hat an i nt egr al quot i ent and a r emai nder are
obt ai ned. Round off t he val ue i n t he fol l owi ng manner :
a) I f t he remai nder i s l ess t han n/2, t he val ue shal l be r ounded down such t hat
t he r ounded val ue i s an i nt egr al mul t i pl e of n.
TABLE I EXAMPLES OF ROUNDI NG OFF VALUES TO UNI T
FI NENESS
VALUE FI NENESS OF ROUNDI NG
1 0.1 0.01 0.001
Rounded
Val ue
Rul e Rounded
Val ue
Rul e Rounded
Val ue
Rul e Rounded
Val ue
Rul e
7.260 4 7 I 7.3 I I 7.26 I 7.260 I
14.725 15 I I 14.7 I 14.72 I I I (b) 14.725 —
3.455 3 I 3.5 I I 3.46 I I I (a) 3.455 —
13.545 001 14 I I 13.5 I 13.55 I I 13.545 I
8.725 9 I I 8.7 I 8.72 I I I (b) 8.725 —
19.205 19 I 19.2 I 19.20 I I I (b) 19.205 —
0.549 9 1 I I 0.5 I 0.55 I I 0.550 I I
0.650 1 1 I I 0.7 I I 0.65 I 0.650 I
0.049 50 0 I 0.0 I 0.05 I I 0.050 I I I (a)
I S : 2  1960
7
b) I f t he r emai nder i s gr eat er t han n/2, t he val ue shal l be r ounded up such t hat
t he r ounded val ue i s an i nt egr al mul t i pl e of n.
c) I f t he r emai nder i s exact l y equal t o n/2, t hat r ounded val ue shal l be chosen
whi ch i s an i nt egr al mul t i pl e of 2n.
3.2.1 Fi neness of r oundi ng ot her t han 2 and 5 i s sel dom cal l ed for i n
pr act i ce. For t hese cases, t he r ul es for r oundi ng may be st at ed i n si mpl er
for m as fol l ows:
a) Roundi ng off t o fi neness 50, 5, 0.5, 0.05, 0.005, et c.
Rul e V — When r oundi ng t o 5 uni t s, t he gi ven val ue shal l be
doubl ed and r ounded off t o t wi ce t he r equi r ed fi neness of r oundi ng i n
accor dance wi t h 3.1.1. The val ue t hus obt ai ned shal l be hal ved t o get
t he fi nal r ounded val ue.
For exampl e, i n r oundi ng off 975 t o t he near est 50, 975 i s doubl ed
gi vi ng 1 950 whi ch becomes 2 000 when r ounded off t o t he near est 100;
when 2 000 i s di vi ded by 2, t he r esul t i ng number 1 000 i s t he r ounded
val ue of 975.
b) Roundi ng off t o fi neness 20, 2, 0.2, 0.02, 0.002, et c.
Rul e VI — When r oundi ng t o 2 uni t s, t he gi ven val ue shal l be
hal ved and r ounded off t o hal f t he r equi r ed fi neness of r oundi ng i n
accor dance wi t h 3.1. The val ue t hus obt ai ned shal l t hen be doubl ed t o
get t he fi nal r ounded val ue.
For exampl e, i n r oundi ng off 2.70 t o t he near est 0.2, 2.70 i s hal ved
gi vi ng 1.35 whi ch becomes 1.4 when r ounded off t o t he near est 0.1;
when 1.4 i s doubl ed, t he r esul t i ng number 2.8 i s t he r ounded val ue.
TABLE I I EXAMPLES OF ROUNDI NG OFF VALUES TO
FI NENESS OTHER THAN UNI T
VALUE FI NENESS OF
ROUNDI NG, n
QUOTI ENT ROUNDED
QUOTI ENT
FI NAL ROUNDED
VALUE
(1) (2) (3) = (1)/(2) (4) (5) = (2) × (4)
1.647 8 0.2 8.239 8 1.6
2.70 0.2 13.5 14 2.8
2.496 8 0.3 8.322 7 8 2.4
1.75 0.5 3.5 4 2.0
0.687 21 0.07 9.817 3 10 0.70
0.875 0.07 12.5 12 0.84
325 50 6.5 6
3 × 10
2
1 025 50 20.5 20
10 ×10
2
I S : 2  1960
8
3.3 Successi ve Rounding — The fi nal r ounded val ue shal l be obt ai ned
fr om t he most pr eci se val ue avai l abl e i n one st ep onl y and not fr om a
ser i es of successi ve r oundi ngs. For exampl e, t he val ue 0.549 9, when
r ounded t o one si gni fi cant fi gur e, shal l be wr i t t en as 0.5 and not as 0.6
whi ch i s obt ai ned as a r esul t of successi ve r oundi ngs t o 0.550, 0.55,
and 0.6. I t i s obvi ous t hat t he most pr eci se val ue avai l abl e i s near er
t o 0.5 and not t o 0.6 and t hat t he er r or i nvol ved i s l ess i n t he for mer
case. Si mi l ar l y, 0.650 1 shal l be r ounded off t o 0.7 i n one st ep and not
successi vel y t o 0.650, 0.65 and 0.6, si nce t he most pr eci se val ue avai l abl e
her e i s near er t o 0.7 t han t o 0.6 ( see al so Tabl e I ).
NOTE 5 — I n t hose cases wher e a fi nal r ounded val ue t er mi nat es wi t h 5 and i t
i s i nt ended t o use i t i n furt her comput at i on, i t may be hel pful t o use a ‘+’ or ‘–’ si gn
aft er t he fi nal 5 t o i ndi cat e whet her a subsequent roundi ng shoul d be up or down.
Thus 3.214 7 may be wr i t t en as 3.215– when r ounded t o a fi neness of
r oundi ng 0.001. I f fur t her r oundi ng t o t hr ee si gni fi cant fi gur es i s desi red, t hi s
number woul d be r ounded down and wr i t t en as 3.21 whi ch i s i n err or by l ess t han
hal f a uni t i n t he l ast pl ace; ot her wi se, r oundi ng of 3.215 woul d have yi el ded 3.22
whi ch i s i n er r or by mor e t han hal f a uni t i n t he l ast pl ace. Si mi l ar l y, 3.205 4 coul d
be wri t t en as 3.205+ when rounded t o 4 si gni fi cant fi gur es. Fur t her r oundi ng t o 3
si gni fi cant fi gur es woul d yi el d t he val ue as 3.21.
I n case t he fi nal 5 i s obt ai ned exact l y, i t woul d be i ndi cat ed by l eavi ng t he 5 as
such wi t hout usi ng ‘+’ or ‘–’ si gn. I n subsequent r oundi ng t he 5 woul d t hen be
t r eat ed i n accor dance wi t h Rul e I I I .
3.4 The r ul es gi ven i n 3.1, 3.2 and 3.3 shoul d be used onl y i f no speci fi c
cr i t er i a for t he sel ect i on of t he r ounded number have t o be t aken i nt o
account . I n al l cases, wher e safet y r equi r ement s or pr escr i bed l i mi t s
have t o be r espect ed, r oundi ng off shoul d be done i n one di r ect i on onl y.
4. NUMBER OF FI GURES TO BE RETAI NED
4.0 Per t i nent t o t he appl i cat i on of t he r ul es for r oundi ng off i s t he
under l yi ng deci si on as t o t he number of fi gur es t hat shoul d be r et ai ned
i n a gi ven pr obl em. The or i gi nal val ues r equi r i ng t o be r ounded off may
ar i se as a r esul t of a t est , an anal ysi s or a measur ement , i n ot her wor ds,
exper i ment al r esul t s, or t hey may ar i se fr om comput at i ons i nvol vi ng
sever al st eps.
4.1 Exper i ment al Result s — The number of fi gur es t o be r et ai ned i n
an exper i ment al r esul t , ei t her for t he pur pose of r epor t i ng or for gui di ng
t he for mul at i on of speci fi cat i ons wi l l depend on t he si gni fi cance of t he
fi gur es i n t he val ue. Thi s aspect has been di scussed i n det ai l under 4 of
I S : 7871956 t o whi ch r efer ence may be made for obt ai ni ng hel pful
gui dance.
4.2 Comput at ions — I n comput at i ons i nvol vi ng val ues of di ffer ent
accur aci es, t he pr obl em as t o how many fi gur es shoul d be r et ai ned at
var i ous st eps assumes a speci al si gni fi cance as i t woul d affect t he
accur acy of t he fi nal r esul t . The r oundi ng off er r or wi l l , i n fact , be
i nject ed i nt o comput at i on ever y t i me an ar i t hmet i cal oper at i on i s
per for med. I t i s, t her efor e, necessar y t o car r y out t he comput at i on i n
I S : 2  1960
9
such a manner as woul d obt ai n accur at e r esul t s consi st ent wi t h t he
accur acy of t he dat a i n hand.
4.2.1 Whi l e i t i s not possi bl e t o pr escr i be det ai l s whi ch may be fol l owed
i n comput at i ons of var i ous t ypes, cer t ai n basi c r ul es may be
r ecommended for si ngl e ar i t hmet i cal oper at i ons whi ch, when fol l owed,
wi l l save l abour and at t he same t i me enabl e accur acy of or i gi nal dat a t o
be nor mal l y mai nt ai ned i n t he fi nal answer s.
4.2.2 As a gui de t o t he number of pl aces or fi gur es t o be r et ai ned i n t he
cal cul at i ons i nvol vi ng ar i t hmet i cal oper at i ons wi t h r ounded or
appr oxi mat e val ues, t he fol l owi ng pr ocedur es ar e r ecommended:
a) Addi ti on — The mor e accur at e val ues shal l be r ounded off so as t o
r et ai n one mor e pl ace t han t he l ast si gni fi cant fi gur e i n t he l east
accur at e val ue. The r esul t i ng sum shal l t hen be r ounded off t o t he
l ast si gni fi cant pl ace i n t he l east accur at e val ue.
b) Subtr acti on — The mor e accur at e val ue (of t he t wo gi ven val ues)
shal l be r ounded off, befor e subt r act i on, t o t he same pl ace as t he
l ast si gni fi cant fi gur e i n l ess accur at e val ue; and t he r esul t shal l be
r epor t ed as such ( see al so Not e 6 ).
c) Mul ti pl i cati on and Di vi si on — The number of si gni fi cant fi gur es
r et ai ned i n t he mor e accur at e val ues shal l be kept one mor e t han
t hat i n t he l east accur at e val ue. The r esul t shal l t hen be r ounded
off t o t he same number of si gni fi cant fi gur es as i n t he l east
accur at e val ue.
d) When a l ong comput at i on i s car r i ed out i n sever al st eps, t he
i nt er medi at e r esul t s shal l be pr oper l y r ounded at t he end of each
st ep so as t o avoi d t he accumul at i on of r oundi ng er r or s i n such
cases. I t i s r ecommended t hat , at t he end of each st ep, one mor e
si gni fi cant fi gur e may be r et ai ned t han i s r equi r ed under (a), (b)
and (c) ( see al so Not e 7 ).
NOTE 6 — The l oss of t he si gni fi cant fi gur es i n t he subt r act i on of t wo near l y
equal val ues i s t he gr eat est sour ce of i naccur acy i n most comput at i ons, and i t
for ms t he weakest l i nk i n a chai n comput at i on where i t occur s. Thus, i f t he val ues
0.169 52 and 0.168 71 ar e each cor r ect t o fi ve si gni fi cant f i gur es, t hei r
di ffer ence 0.000 81, whi ch has onl y t wo si gni fi cant fi gur es, i s qui t e l i kel y t o
i nt r oduce i naccur acy i n subsequent comput at i on.
I f, however, t he di ffer ence of t wo val ues i s desi r ed t o be corr ect t o k si gni fi cant
fi gur es and i f i t i s known befor ehand t hat t he fi r st m si gni fi cant fi gures at t he l eft
wi l l di sappear by subt r act i on, t hen t he number of si gni fi cant fi gur es t o be r et ai ned
i n each of t he val ues shal l be m + k ( see Exampl e 4 ).
NOTE 7 — To ensur e a gr eat er degr ee of accur acy i n t he comput at i ons, i t i s al so
desi r abl e t o avoi d or def er as l ong as possi bl e cer t ai n appr oxi mat i on oper at i ons
l i ke t hat of t he di vi si on or squar e root . For exampl e, i n t he det er mi nat i on of
sucr ose by vol umet ri c met hod, t he expr essi on may be bet t er
eval uat ed by t aki ng i t s cal cul at i onal for m as 20w
1
( f
2
v
1
– f
1
v
2
)/w
2
v
1
v
2
whi ch
woul d def er t he di vi si on unt i l t he l ast oper at i on of t he cal cul at i on.
20w
1
w
2

f
2
v
2

f
1
v
1

\ .

 
I S : 2  1960
10
4.2.3 Exampl es
Exampl e 1
Requi r ed t o fi nd t he sum of t he r ounded off val ues 461.32, 381.6,
76.854 and 4.746 2.
Si nce t he l east accur at e val ue 381.6 i s known onl y t o t he fi r st deci mal
pl ace, al l ot her val ues shal l be r ounded off t o one mor e pl ace, t hat
i s, t o t wo deci mal pl aces and t hen added as shown bel ow:
The r esul t i ng sum shal l t hen be r epor t ed t o t he same deci mal pl ace as
i n t he l east accur at e val ue, t hat i s, as 924.5.
Exampl e 2
Requi r ed t o fi nd t he sum of t he val ues 28 490, 894, 657.32, 39 500
and 76 939, assumi ng t hat t he val ue 39 500 i s known t o t he
near est hundr ed onl y.
Si nce one of t he val ues i s known onl y t o t he near est hundr ed, t he
ot her val ues shal l be r ounded off t o t he near est t en and t hen added
as shown bel ow:
The sum shal l t hen be r epor t ed t o t he near est hundr ed as 1 465 × 100
or even as 1.465 × 10
5
.
Exampl e 3
Requi r ed t o fi nd t he di ffer ence of 679.8 and 76.365, assumi ng t hat
each number i s known t o i t s l ast fi gur e but no far t her .
461.32
381.6
76.85
4.75
924.52
2 849 × 10
89 × 10
66 × 10
3 950 × 10
7 694 × 10
14 648 × 10
I S : 2  1960
11
Si nce one of t he val ues i s known t o t he fi r st deci mal pl ace onl y, t he
ot her val ue shal l al so be r ounded off t o t he fi r st deci mal pl ace and
t hen t he di ffer ence shal l be found.
The di ffer ence, 603.4, shal l be r epor t ed as such.
Exampl e 4
Requi r ed t o eval uat e cor r ect t o fi ve si gni fi cant fi gur es.
Si nce = 1.587 450 79
= 1.577 973 38
and t hr ee si gni fi cant fi gur es at t he l eft wi l l di sappear on
subt r act i on, t he number of si gni fi cant fi gur es r et ai ned i n each
val ue shal l be 8 as shown bel ow:
The r esul t , 0.009 477 4, shal l be r epor t ed as such (or
as 9.477 4 × 10
–3
).
Exampl e 5
Requi r ed t o eval uat e 35.2/ gi ven t hat t he numer at or i s cor r ect t o
i t s l ast fi gur e.
Si nce t he numer at or her e i s cor r ect t o t hr ee si gni fi cant fi gur es, t he
denomi nat or shal l be t aken as = 1.414. Then,
= 24.89
and t he r esul t shal l be r epor t ed as 24.9.
Exampl e 6
Requi r ed t o eval uat e 3.78t/5.6, assumi ng t hat t he denomi nat or i s
t r ue t o onl y t wo si gni fi cant fi gur es.
Si nce t he denomi nat or her e i s cor r ect t o t wo si gni fi cant fi gur es, each
number i n t he numer at or woul d be t aken up t o t hr ee si gni fi cant
679.8
76.4
603.4
1.587 450 8
1.577 973 4
0.009 477 4
2.52 2.49
2.52
2.49
2,
2
35.2
1.414

I S : 2  1960
12
fi gur es. Thus,
= 2,08.
The r esul t shal l , however , be r epor t ed as 2.1.
A P P E N D I X A
( Cl ause3.0 )
VALI DI TY OF RULES
A1. The val i di t y of t he r ul es for r oundi ng off numer i cal val ues, as gi ven
i n 3.1, may be seen fr om t he fact t hat t o ever y number t hat i s t o be
‘r ounded down’ i n accor dance wi t h Rul e I , t her e cor r esponds a number
t hat i s t o be ‘r ounded up’ i n accor dance wi t h Rul e I I . Thus, t hese t wo
r ul es est abl i sh a bal ance bet ween r oundi ng ‘down’ and ‘up’ for al l
number s ot her t han t hose t hat fal l exact l y mi dway bet ween t wo
al t er nat i ves. I n t he l at t er case, si nce t he fi gur e t o be dr opped i s
exact l y 5, Rul e I I I , whi ch speci fi es t hat t he val ue shoul d be r ounded t o
i t s near est even number , i mpl i es t hat r oundi ng shal l be ‘up’ when t he
pr ecedi ng fi gur es ar e 1, 3, 5, 7, 9 and ‘down’ when t hey ar e 0, 2, 4, 6, 8.
Rul e I I I hence advocat es a si mi l ar bal ance bet ween r oundi ng ‘up’ and
‘down’ ( see al so Not e 8 ). Thi s i mpl i es t hat i f t he above r ul es ar e fol l owed
i n a l ar ge gr oup of val ues i n whi ch r andom di st r i but i on of fi gur es occur s,
t he number ‘r ounded up’ and t he number ‘r ounded down’ wi l l be near l y
equal . Ther efor e, t he sum and t he aver age of t he r ounded val ues wi l l be
mor e near l y cor r ect t han woul d be t he case i f al l wer e r ounded i n t he
same di r ect i on, t hat i s, ei t her al l ‘up’ or al l ‘down’.
NOTE 8 — Fr om pur el y l ogi cal consi der at i ons, a gi ven val ue coul d have as wel l
been r ounded t o an odd number (and not an even number as i n Rul e I I I ) when t he
di scar ded fi gur es fal l exact l y mi dway bet ween t wo al t er nat i ves. But t her e i s a
pr act i cal aspect t o t he mat t er . The roundi ng off a val ue t o an even number
faci l i t at es t he di vi si on of t he r ounded val ue by 2 and t he r esul t of such di vi si on
gi ves t he corr ect r oundi ng of hal f t he or i gi nal unr ounded val ue. Besi des, t he
(r ounded) even val ues may gener al l y be exact l y di vi si bl e by many mor e number s,
even as wel l as odd, t han ar e t he (rounded) odd val ues.
3.78 3.14 ×
5.7

Bur eau of I ndi an Standar ds
BI S i s a st at ut or y i nst i t ut i on est abl i shed under t he Bur eau of I ndi an Standar ds Act , 1986 t o
pr omot e har moni ous devel opment of t he act i vi t i es of st andar di zat i on, mar ki ng and qual i t y
cer t i fi cat i on of goods and at t endi ng t o connect ed mat t er s i n t he count r y.
Copyr ight
BI S has t he copyr i ght of al l i t s publ i cat i ons. No par t of t hese publ i cat i ons may be r epr oduced i n
any for m wi t hout t he pr i or per mi ssi on i n wr i t i ng of BI S. Thi s does not pr ecl ude t he fr ee use, i n t he
cour se of i mpl ement i ng t he st andar d, of necessar y det ai l s, such as symbol s and si zes, t ype or gr ade
desi gnat i ons. Enqui r i es r el at i ng t o copyr i ght be addr essed t o t he Di r ect or (Publ i cat i ons), BI S.
Revi ew of I ndi an St andar ds
Amendment s ar e i ssued t o st andar ds as t he need ar i ses on t he basi s of comment s. St andar ds ar e
al so r evi ewed per i odi cal l y; a st andar d al ong wi t h amendment s i s r eaffi r med when such r evi ew
i ndi cat es t hat no changes ar e needed; i f t he r evi ew i ndi cat es t hat changes ar e needed, i t i s t aken
up for revi si on. User s of I ndi an St andar ds shoul d ascer t ai n t hat t hey ar e i n possessi on of t he l at est
amendment s or edi t i on by r efer r i ng t o t he l at est i ssue of ‘BI S Cat al ogue’ and ‘St andar ds : Mont hl y
Addi t i ons’.
Thi s I ndi an St andar d has been devel oped by Techni cal Commi t t ee : EDC 1
Amendments I ssued Si nce Publi cati on
Amend No. Date of I ssue
Amd. No. 1 Febur ar y 1997
Amd. No. 2 Oct ober 1997
Amd. No. 3 August 2000
BUREAU OF I NDI AN STANDARDS
Headquar t ers:
Manak Bhavan, 9 Bahadur Shah Zafar Mar g, New Del hi 110002.
Tel ephones: 323 01 31, 323 33 75, 323 94 02
Tel egr ams: Manaksanst ha
(Common t o al l offi ces)
Regi onal Offi ces: Tel ephone
Cent r al : Manak Bhavan, 9 Bahadur Shah Zafar Mar g
NEW DELHI 110002
323 76 17
323 38 41
East er n : 1/14 C. I . T. Scheme VI I M, V. I . P. Road, Kankur gachi
KOLKATA 700054
337 84 99, 337 85 61
337 86 26, 337 91 20
Nor t her n : SCO 335336, Sect or 34A, CHANDI GARH 160022 60 38 43
60 20 25
Sout her n : C. I . T. Campus, I V Cr oss Road, CHENNAI 600113
235 02 16, 235 04 42
235 15 19, 235 23 15
West er n : Manakal aya, E9 MI DC, Marol , Andher i (East )
MUMBAI 400093
832 92 95, 832 78 58
832 78 91, 832 78 92
Br anches : AH M EDABAD. BANGAL ORE. BH OPAL . BH U BAN ESH WAR.
COI M BATORE. F ARI DABAD. GH AZI ABAD. GU WAH AT I .
H YDERABAD. J AI PU R. K AN PU R. L UCK N OW. N AGPUR.
N AL AGARH . PATN A. PU N E. RAJ K OT . T H I RUVAN AN TH APU RAM.
VI SH AK H APAT NAM .
¹
´
¦
¹
´
¦
¹
´
¦
¹
´
¦
¹
´
¦
Design & Standardization Organization (Ministry of Railways) SHRI B ALESHWAR N ATH Central Board of Irrigation & Power. R. KRISHNAN Indian Institute of Science. S. S. New Delhi BRIG J. N IJHAWAN Council of Scientific & Industrial Research. R. Calcutta DR R. BHATTACHARJEE ( Alternate ) DR B. JOSHI Institution of Engineers (India). B. 9 BAHADUR SHAH ZAFAR MARG NEW DELHI 110002 . MEHRA Council of Scientific & Industrial Research. N. S. N. Bangalore SHRI S. Calcutta SHRI K. KRISHNAN Members SHRI PREM PRAKASH ( Alternate to Dr K. New Delhi SHRI S. New Delhi DIRECTOR Engineering Research Department. Krishnan ) ASSISTANT DIRECTOR Research. D. KAPUR Indian Engineering Association. SAMSON Controller of Development (Armaments) (Ministry LTCOL R. New Delhi DEPUTY DIRECTOR GENERAL OF Directorate General of Observatories (Ministry of OBSERVATORIES Transport & Communications).1960 Indian Standard RULES FOR ROUNDING OFF NUMERICAL VALUES ( Revised ) Engineering Standards Sectional Committee. EDC 1 Chairman DR K. R. Government of Andhra Pradesh SHRI S. R AGHAVAN Central Water & Power Commission. Calcutta SHRI R. MUKERJI Government Test House. New Delhi BUREAU OF INDIAN STANDARDS MANAK BHAVAN . R. JANARDHANAM of Defence) ( Alternate ) ( Continued on page 2 ) Council of Scientific & Industrial Research.IS : 2 . New Delhi SHRI V.
SINHA SHRI J. ISI 2 . K. TREHAN SHRI T. N. M. N. VERMAN ( Exofficio ) SHRI J. N. ISI Extra Assistant Director (Stat). SINGH Assistant Director (Stat). MEHROTRA ( Alternate ) Secretaries DR A. SARMA SHRI J. Housing & Supply) Engineering Association of India. GUPTA SHRI B. ISI Directorate General of Supplies & Disposals (Ministry of Works. M. New Delhi Director. BHARGAVA ( Alternate ) LTGEN H.1960 ( Continued from page 1 ) Members SHRI R. ISI Deputy Director (Engg). Calcutta Ministry of Transport & Communications (Roads Wing) Council of Scientific & Industrial Research. P. W ILLIAMS DR LAL C.IS : 2 .
it is necessary to prescribe rules for ‘rounding off’ numerical values as also for deciding on ‘the number of figures’ to be retained. Moreover.2 To round off a value is to retain a certain number of figures.4 In preparing this standard. British Standards Institution. F O R E W O R D 0. AMERICAN STANDARD Z 25. 0.11940 RULES FOR ROUNDING OFF NUMERICAL VALUES. the rules given referred only to unit fineness of rounding ( see 2.1960 Indian Standard RULES FOR ROUNDING OFF NUMERICAL VALUES ( Revised ) 0. ROUNDING OF NUMBERS). However. reference has been made to the following: IS : 7871956 GUIDE FOR INTERCONVERSION OF VALUES FROM ONE SYSTEM OF UNITS TO ANOTHER. counted from the left.3 ) and in course of years the need was felt to prescribe rules for rounding off numerical values to fineness of rounding other than unity.IS : 2 . and drop the others so as to give a more rational form to the value. American Standards Association. after the draft finalized by the Engineering Standards Sectional Committee had been approved by the Engineering Division Council. BS 1957 : 1953 PRESENTATION OF NUMERICAL VALUES (FINENESS OF EXPRESSION. The present revision is expected to fulfil these needs.3 This standard was originally issued in 1949 with a view to promoting the adoption of a uniform procedure in rounding off numerical values. IS : 1890 (PART 0)1995/ISO 310 : 1992 QUANTITIES AND UNITS : PART 0 GENERAL PRINCIPLES ( FIRST REVISION ). Numerical Mathematical Analysis Baltimore. 1955. As the result of a test or of a calculation is generally rounded off for the purpose of reporting or for drafting specifications. it was also felt that the discussion on the number of figures to be retained as given in the earlier version required further elucidation. 3 . JAMES W. 0. S CARBOROUGH. American Society for Testing and Materials. The John Hopkins Press. ASTM DESIGNATION : E 2950 RECOMMENDED PRACTICE FOR DESIGNATING SIGNIFICANT PLACES IN SPECIFIED VALUES. Indian Standards Institution.1 This Indian Standard (Revised) was adopted by the Indian Standards Institution on 27 July 1960. 0.
029 5 2 000. 1 (Feburary 1997).0 For the purpose of this standard. counting from the first figure after the decimal point and ending with the last figure on the right. Examples: Value 0. Amendment No. 2 (October 1997) and Amendment No. a measurement or a calculation.000 001 5 677.029 50 21. and thus assisting in drafting specifications. the following definitions shall apply. 1.1 Number of Decimal Places — A value is said to have as many decimal places as there are number of figures in the value. It also makes recommendations as to the number of figures that should be retained in course of computation.029 5 10.3 incorporates Amendment No. an analysis. TERMINOLOGY 2. 2. the expression 10. counting from the leftmost nonzero digit and ending with the rightmost digit in the value. 10 3.32 × 103 ( see Note 1 ) Decimal Places 5 4 6 2 2 NOTE 1 — For the purpose of this standard.77 × 102 0 056. 2.1960 0. SCOPE 1.1 This standard prescribes rules for rounding off numerical values for the purpose of reporting results of a test.IS : 2 .32 and the unit of expression for the value. 2.029 500 0. Side bar indicates modification of the text as the result of incorporation of the amendments.029 5 2 000. Examples: Value 0. 3 (August 2000).00 10.2 Number of Significant Figures — A value is said to have as many significant figures as there are number of significant digits ( see Note 2 ) in the value. the value proper which is 10.5 This edition 2.32 × 103 should be taken to consist of two parts.0 567 700 56.770 3 900 ( see Note 3 ) Significant Figures 5 3 6 10 5 6 4 5 4 4 .000 001 291.
. In such cases. 3. 50. For example. For example. Obviously. the sum and the average of a series of values so rounded will be larger or smaller than the corresponding sum or average of the unrounded values.1 in one step ( see 3.3 Fineness of Rounding — The unit to which a value is rounded off.5.1 Rounding Off to Unit Fineness — In case the fineness of rounding is unity in the last place retained. 0. When appearing in the power of 10 to indicate the magnitude of the unit in the expression of a value. the figure in the last place retained shall be increased by 1.4) shall be followed: Rule I — When the figure next beyond the last figure or place to be retained is less than 5. RULES FOR ROUNDING 3. 3 900 may be written as 3.90 × 103 or 3. 3.. that is. 10. increase by 1. However. discard everything beyond the last figure retained.0 The rule usually followed in rounding off a value to unit fineness of rounding is to keep unchanged the last figure retained when the figure next beyond is less than 5 and to increase by 1 the last figure retained when the figure next beyond is more than 5.001. 0.... 20. 2. zero shall not be a significant digit. 5 .07.000 5. 2.. if rounding off is carried out in accordance with the rules stated in 3. some computers ‘round up’.000 01. it would be always desirable to write the value in the poweroften notation. Rule III — When the figure next beyond the last figure or place to be retained is 5 alone or 5 followed by zeros only. 100 or any other unit depending on the fineness desired. the figure in the last place retained shall be (a) increased by 1 if it is odd and (b) left unchanged if even (zero would be regarded as an even number for this purpose). 3.01. 0. others ‘round down’.9 × 103. 2. 1. Rule II — When the figure next beyond the last figure or place to be retained is more than 5 or is 5 followed by any figures other than zeros. NOTE 3 — With a view to removing any ambiguity regarding the significance of the zeros at the end in a value like 3 900.IS : 2 .. There is diversity of practice when the figure next beyond the last figure retained is 5.9 occurring in a value shall be a significant digit(s).. 0. 1. 0. the figure in the last place retained shall be left unchanged. a value may be rounded to the nearest 0.900 × 103 depending upon the last figure(s) in the value to which it is desired to impart significance. 0.. if the retained value is always ‘rounded up’ or always ‘rounded down’.000 2... the following rules (except in 3.1960 NOTE 2 — Any of the digits. 0.3 ). the last figure retained.002 5.005. and zero shall be a significant digit only when it is preceded by some other digit (excepting zeros) on its left. that is. the sum and the average of the rounded values would be more nearly correct than in the previous cases ( see Appendix A ). 3..
2.55 0.0 Rule II I II I I I I II I OF ROUNDING 0. 100. Some examples illustrating the application of Rule IV are given in Table II.40. Round off the value in the following manner: a) If the remainder is less than n/2.455 13.650 0.65 0.26 14.1.455 13.1960 Some examples illustrating the application of Rules I to III are given in Table I. N OTE 4 — The rules for rounding off a value to any fineness of rounding.545 001 8.050 7.1 Rounded Value 7.1 for unit fineness of rounding.55 8. For example.3 14.7 3. other than unity.46 13. Similarly.650 1 0. say.205 3.725 III(b) 19. 10.001 Rounded Rule Value 7. 6 .01 Rounded Value 7.549 9 0.205 0. the value shall be rounded down such that the rounded value is an integral multiple of n. The number so obtained.72 3. shall then be multiplied by n to get the final rounded value. 712 and 715 when rounded to the fineness 10 become 710 and 720 respectively.5 0.2 Rounding Off to Fineness Other than Unity — In case the fineness of rounding is not unity. n.2 0. 1 000.260 I — — I — — II I III(a) 3.20 0.5 13. the given value shall be rounded off according to the following rule: Rule IV — When rounding to a fineness n. etc. 3.550 0.049 50 III(b) 14.10.43 when rounded to fineness 0. that is the rounded quotient.10 becomes 2.725 19. it is n. TABLE I VALUE 1 Rounded Rule Value 7 I 15 3 14 9 19 1 1 0 II I II II I II II I EXAMPLES OF ROUNDING OFF VALUES TO UNIT FINENESS FINENESS 0.725 0.72 19.1 may be extended to apply when the fineness of rounding is 0.05 Rule I III(a) II III(b) II I II 0.545 8.1 The rules for rounding laid down in 3. The quotient shall be rounded off to the nearest whole number in accordance with the rules laid down in 3.IS : 2 .260 4 14.7 19.1 ) as follows: Divide the given value by n so that an integral quotient and a remainder are obtained. may also be stated in line with those for unit fineness of rounding ( see 3. the given value shall be divided by n. but.5 8.7 0.725 3.
n (2) 0. the resulting number 1 000 is the rounded value of 975. in rounding off 975 to the nearest 50. the rules for rounding may be stated in simpler form as follows: a) Rounding off to fineness 50. etc. 2.5 6.647 8 2. 5.817 3 12. 0.70 to the nearest 0.02.5 ROUNDED QUOTIENT (4) 8 14 8 4 10 12 6 20 FINAL ROUNDED VALUE (5) = (2) × (4) 1.496 8 1.1.5. 0.5 8.322 7 3. the given value shall be doubled and rounded off to twice the required fineness of rounding in accordance with 3. For example.1. 0. 0. etc.6 2. 2. 0. c) If the remainder is exactly equal to n/2. the resulting number 2.70 0.IS : 2 . For these cases.4 when rounded off to the nearest 0.2.1960 b) If the remainder is greater than n/2.4 is doubled.875 325 1 025 EXAMPLES OF ROUNDING OFF VALUES TO FINENESS OTHER THAN UNIT QUOTIENT (3) = (1)/(2) 8. TABLE II VALUE (1) 1.84 3 × 102 10 ×102 FINENESS OF ROUNDING. when 2 000 is divided by 2. the given value shall be halved and rounded off to half the required fineness of rounding in accordance with 3.687 21 0.07 50 50 3.0 0.70 2. the value shall be rounded up such that the rounded value is an integral multiple of n. The value thus obtained shall be halved to get the final rounded value. 0.4 2. that rounded value shall be chosen which is an integral multiple of 2n.1 Fineness of rounding other than 2 and 5 is seldom called for in practice.5 0. Rule V — When rounding to 5 units.8 2. 7 . b) Rounding off to fineness 20. The value thus obtained shall then be doubled to get the final rounded value. in rounding off 2.3 0.005. Rule VI — When rounding to 2 units.5 9.2 0.2.1.2 0.8 is the rounded value.239 13.5 20.2.002. 975 is doubled giving 1 950 which becomes 2 000 when rounded off to the nearest 100.07 0.05.75 0.1. when 1.70 is halved giving 1.35 which becomes 1. For example.
this number would be rounded down and written as 3.205+ when rounded to 4 significant figures. and 0.215 would have yielded 3.001.4 The rules given in 3.215– when rounded to a fineness of rounding 0. It is obvious that the most precise value available is nearer to 0.21 which is in error by less than half a unit in the last place. 3. where safety requirements or prescribed limits have to be respected. the problem as to how many figures should be retained at various steps assumes a special significance as it would affect the accuracy of the final result.22 which is in error by more than half a unit in the last place. In all cases.6 and that the error involved is less in the former case. This aspect has been discussed in detail under 4 of IS : 7871956 to which reference may be made for obtaining helpful guidance.6.549 9. or they may arise from computations involving several steps. Further rounding to 3 significant figures would yield the value as 3. 0. in other words. In subsequent rounding the 5 would then be treated in accordance with Rule III. 0. NUMBER OF FIGURES TO BE RETAINED 4.3 should be used only if no specific criteria for the selection of the rounded number have to be taken into account. rounding off should be done in one direction only.1. The original values requiring to be rounded off may arise as a result of a test. it may be helpful to use a ‘+’ or ‘–’ sign after the final 5 to indicate whether a subsequent rounding should be up or down. otherwise.7 in one step and not successively to 0. be injected into computation every time an arithmetical operation is performed. it would be indicated by leaving the 5 as such without using ‘+’ or ‘–’ sign.6 which is obtained as a result of successive roundings to 0. in fact. when rounded to one significant figure.7 than to 0. 4.214 7 may be written as 3. Similarly. 0.5 and not as 0. In case the final 5 is obtained exactly. It is. 3.IS : 2 .6.2 and 3. Similarly. necessary to carry out the computation in 8 .650 1 shall be rounded off to 0. an analysis or a measurement.55. therefore.65 and 0. either for the purpose of reporting or for guiding the formulation of specifications will depend on the significance of the figures in the value. 4.6 ( see also Table I ). the value 0.3 Successive Rounding — The final rounded value shall be obtained from the most precise value available in one step only and not from a series of successive roundings.550.1960 3. since the most precise value available here is nearer to 0.205 4 could be written as 3.0 Pertinent to the application of the rules for rounding off is the underlying decision as to the number of figures that should be retained in a given problem. Thus 3. If further rounding to three significant figures is desired.650. experimental results. 4.21. 3.5 and not to 0. N OTE 5 — In those cases where a final rounded value terminates with 5 and it is intended to use it in further computation.1 Experimental Results — The number of figures to be retained in an experimental result.2 Computations — In computations involving values of different accuracies. shall be written as 0. For example. rounding of 3. The rounding off error will.
if the values 0. The resulting sum shall then be rounded off to the last significant place in the least accurate value. it is also desirable to avoid or defer as long as possible certain approximation operations like that of the division or square root. and it forms the weakest link in a chain computation where it occurs. when followed. The result shall then be rounded off to the same number of significant figures as in the least accurate value. b) Subtraction — The more accurate value (of the two given values) shall be rounded off. to the same place as the last significant figure in less accurate value. 9 . before subtraction.IS : 2 . one more significant figure may be retained than is required under (a). 4. N OTE 7 — To ensure a greater degree of accuracy in the computations. which has only two significant figures. the difference of two values is desired to be correct to k significant figures and if it is known beforehand that the first m significant figures at the left will disappear by subtraction. For example. is quite likely to introduce inaccuracy in subsequent computation. 4.1960 such a manner as would obtain accurate results consistent with the accuracy of the data in hand. then the number of significant figures to be retained in each of the values shall be m + k ( see Example 4 ). d) When a long computation is carried out in several steps.168 71 are each correct to five significant figures. It is recommended that. c) Multiplication and Division — The number of significant figures retained in the more accurate values shall be kept one more than that in the least accurate value. however.169 52 and 0. their difference 0. and the result shall be reported as such ( see also Note 6 ).2. If.  may be better v2 v 1 evaluated by taking its calculational form as 20w1 ( f2 v1 – f1 v2 )/w2 v1 v2 which would defer the division until the last operation of the calculation.000 81.2 As a guide to the number of places or figures to be retained in the calculations involving arithmetical operations with rounded or approximate values.2. (b) and (c) ( see also Note 7 ). will save labour and at the same time enable accuracy of original data to be normally maintained in the final answers.1 While it is not possible to prescribe details which may be followed in computations of various types. in the determination of 20w 1 sucrose by volumetric method. at the end of each step. the following procedures are recommended: a) Addition — The more accurate values shall be rounded off so as to retain one more place than the last significant figure in the least accurate value. Thus. the intermediate results shall be properly rounded at the end of each step so as to avoid the accumulation of rounding errors in such cases. certain basic rules may be recommended for single arithmetical operations which. the expression w2 f2 f 1 . NOTE 6 — The loss of the significant figures in the subtraction of two nearly equal values is the greatest source of inaccuracy in most computations.
Since the least accurate value 381.6 is known only to the first decimal place. that is.32. the other values shall be rounded off to the nearest ten and then added as shown below: 2 849 × 10 89 × 10 66 × 10 3 950 × 10 7 694 × 10 14 648 × 10 The sum shall then be reported to the nearest hundred as 1 465 × 100 or even as 1.746 2. 10 .1960 4. Example 2 Required to find the sum of the values 28 490. 381.75 924.32.8 and 76.3 Examples Example 1 Required to find the sum of the rounded off values 461.365. to two decimal places and then added as shown below: 461. 894. as 924.52 The resulting sum shall then be reported to the same decimal place as in the least accurate value. 657.32 381. Since one of the values is known only to the nearest hundred.6.465 × 105.85 4. 76.2. assuming that each number is known to its last figure but no farther.854 and 4. all other values shall be rounded off to one more place.IS : 2 .6 76. assuming that the value 39 500 is known to the nearest hundred only. that is.5. 39 500 and 76 939. Example 3 Required to find the difference of 679.
Since the denominator here is correct to two significant figures.6.477 4 × 10–3).414 and the result shall be reported as 24. 0.4.4 The difference.89 1.= 24. 603.414.577 973 4 0.587 450 8 1.1960 Since one of the values is known to the first decimal place only.52 = 1. shall be reported as such. given that the numerator is correct to its last figure. Since 2.009 477 4. Example 4 Required to evaluate 2.9. Example 5 Required to evaluate 35.8 76.49 correct to five significant figures.IS : 2 . 679. 35.009 477 4 The result. as 9. the number of significant figures retained in each value shall be 8 as shown below: 1. each number in the numerator would be taken up to three significant shall be reported as such (or 11 . Example 6 Required to evaluate 3. the denominator shall be taken as 2 = 1. Since the numerator here is correct to three significant figures.2/ 2.49 = 1.4 603.78/5. the other value shall also be rounded off to the first decimal place and then the difference shall be found.52 2. Then.577 973 38 and three significant figures at the left will disappear on subtraction.587 450 79 2.2 . assuming that the denominator is true to only two significant figures.
4. implies that rounding shall be ‘up’ when the preceding figures are 1. there corresponds a number that is to be ‘rounded up’ in accordance with Rule II. Rule III.08. 2. the (rounded) even values may generally be exactly divisible by many more numbers.1960 figures. 3. Besides. In the latter case. NOTE 8 — From purely logical considerations. that is. 5.78 3. Thus. may be seen from the fact that to every number that is to be ‘rounded down’ in accordance with Rule I. 12 .= 2. Therefore. 9 and ‘down’ when they are 0. the number ‘rounded up’ and the number ‘rounded down’ will be nearly equal. 8. be reported as 2.IS : 2 . 6. either all ‘up’ or all ‘down’. these two rules establish a balance between rounding ‘down’ and ‘up’ for all numbers other than those that fall exactly midway between two alternatives.14 . 7. a given value could have as well been rounded to an odd number (and not an even number as in Rule III) when the discarded figures fall exactly midway between two alternatives.1. This implies that if the above rules are followed in a large group of values in which random distribution of figures occurs. since the figure to be dropped is exactly 5. even as well as odd. however. The validity of the rules for rounding off numerical values. as given in 3. the sum and the average of the rounded values will be more nearly correct than would be the case if all were rounded in the same direction. But there is a practical aspect to the matter. The rounding off a value to an even number facilitates the division of the rounded value by 2 and the result of such division gives the correct rounding of half the original unrounded value.7 The result shall. 5.0 ) A VALIDITY OF RULES A1. Rule III hence advocates a similar balance between rounding ‘up’ and ‘down’ ( see also Note 8 ). APPENDIX ( Clause 3.1. than are the (rounded) odd values. which specifies that the value should be rounded to its nearest even number. Thus. 3.
E9 MIDC. G HAZIABAD . type or grade designations. T. I. 1 Amd. 9 Bahadur Shah Zafar Marg NEW DELHI 110002 Eastern : 1/14 C. Andheri (East) MUMBAI 400093 Branches : A H M E D A B A D . 337 85 61 337 86 26. of necessary details. No. marking and quality certification of goods and attending to connected matters in the country. Road. F A RI D A B A D. Standards are also reviewed periodically. This Indian Standard has been developed by Technical Committee : EDC 1 Amendments Issued Since Publication Date of Issue Feburary 1997 October 1997 August 2000 Amend No. KANPUR. 9 Bahadur Shah Zafar Marg. Marol. if the review indicates that changes are needed. Campus. 832 78 92 Northern : SCO 335336. JAIPUR. 323 94 02 Regional Offices: Central : Manak Bhavan. 337 91 20 60 38 43 60 20 25 235 02 16. VISHAK HAPAT NAM. V. 832 78 58 832 78 91. Telephones: 323 01 31. a standard along with amendments is reaffirmed when such review indicates that no changes are needed. L UCK N OW . No. . T. 2 Amd. such as symbols and sizes. 3 BUREAU OF INDIAN STANDARDS Headquarters: Manak Bhavan. New Delhi 110002. I. CHENNAI 600113 Western : Manakalaya. BIS. Amd. T HI R UV A N A N THA PU R A M. N AGP UR. BHOPAL. Users of Indian Standards should ascertain that they are in possession of the latest amendments or edition by referring to the latest issue of ‘BIS Catalogue’ and ‘Standards : Monthly Additions’. PU N E . CHANDIGARH 160022 Southern : C. in the course of implementing the standard.Bureau of Indian Standards BIS is a statutory institution established under the Bureau of Indian Standards Act. Copyright BIS has the copyright of all its publications. Scheme VII M. I. HY DERABAD. P A TN A. No. No part of these publications may be reproduced in any form without the prior permission in writing of BIS. Sector 34A. 1986 to promote harmonious development of the activities of standardization. N A LA G AR H . This does not preclude the free use. R A JK OT . Kankurgachi KOLKATA 700054 Telegrams: Manaksanstha (Common to all offices) Telephone 323 76 17 323 38 41 337 84 99. IV Cross Road. P. C O IM B A TO R E. 235 04 42 235 15 19. Review of Indian Standards Amendments are issued to standards as the need arises on the basis of comments. G U WAHAT I. 235 23 15 832 92 95. it is taken up for revision. 323 33 75. Enquiries relating to copyright be addressed to the Director (Publications). BANG ALO RE. B HU BA N E SH WA R .
This action might not be possible to undo. Are you sure you want to continue?