Approximation

Approximation (Rounding Off): The mathematical process of replacing a quantity's exact value with a nearby, convenient, and simplified value is called approximation.

Mathematical Symbol: $\approx$ (Read as: "approximately equal to").

1. Underline the digit at the specified rounding position.
2. Inspect the decider digit immediately to its right.
3. Apply the halfway rule ($+1$ if $\ge 5$, maintain if $< 5$).
4. For whole numbers, replace all digits to the right with zeros ($0$); for decimal values, discard subsequent digits.

Example 1: Round $74$ to the nearest ten.
Units digit is $4 < 5$. Tens digit remains $7$. Hence, $74 \approx 70$.

Example 2: Round $78$ to the nearest ten.
Units digit is $8 \ge 5$. Tens digit becomes $7+1=8$. Hence, $78 \approx 80$.

Cascading Rounding Fallacy: Rounding from far right digits backward (e.g. turning $3.646$ to $3.65$ then to $3.7$) is mathematically incorrect. Rounding depends strictly on the single immediate right-hand decider digit.

Retail & Cash Transactions: Point-of-sale systems round total fractional bills $\ge 50$ paise up to the nearest rupee to prevent denomination friction.

Nearest 10: Look at the units digit ($\ge 5 \implies$ tens $+1$, replace units with $0$).

Nearest 100: Look at the tens digit ($\ge 5 \implies$ hundreds $+1$, replace tens and units with $00$).

Nearest 1000: Look at the hundreds digit ($\ge 5 \implies$ thousands $+1$, replace lower digits with $000$).

For the number $4,675$:
To nearest 10: Units digit is $5 \implies \mathbf{4,680}$
To nearest 100: Tens digit is $7 \ge 5 \implies \mathbf{4,700}$
To nearest 1000: Hundreds digit is $6 \ge 5 \implies \mathbf{5,000}$

Problem: Round $83,452$ to (a) nearest 100, (b) nearest 1000.
Solution:
(a) Nearest 100: Tens digit is $5 \implies 83,452 \approx \mathbf{83,500}$.
(b) Nearest 1000: Hundreds digit is $4 < 5 \implies 83,452 \approx \mathbf{83,000}$.

Omitting Placeholder Zeros: Never write $835$ when rounding $83,452$ to the nearest 100. Place-value zeros must be preserved ($83,500$).

National Demographics & Census: Government statistical bureaus round country populations to nearest thousand or million for demographic modeling.

1st decimal place = Tenths (check 2nd place/hundredths)
2nd decimal place = Hundredths (check 3rd place/thousandths)
3rd decimal place = Thousandths (check 4th place/ten-thousandths)

In approximation, $3.5$ and $3.50$ have different experimental significance! If rounding to 2 decimal places yields $3.5$, write $3.50$. The trailing zero certifies precision to the hundredths place.

Given $12.6748$:
To 1 decimal place: 2nd digit is $7 \ge 5 \implies \mathbf{12.7}$
To 2 decimal places: 3rd digit is $4 < 5 \implies \mathbf{12.67}$
To 3 decimal places: 4th digit is $8 \ge 5 \implies \mathbf{12.675}$

Carry Over on 9: Rounding $4.96$ to 1 decimal place results in $9+1=10$, carrying over to give $5.0$, never $4.10$!

Pharmaceutical Chemistry: Drug dosages require 3 decimal place accuracy (milligrams) to prevent patient toxicity.

Fractions like $1/3 = 0.333\dots$, $2/3 = 0.666\dots$, and $22/7 = 3.1428\dots$ repeat infinitely. For practical computation, they must be rounded to a defined number of decimal places.

Rule: Always carry out long division to one decimal place beyond ($N+1$) the required rounding precision $N$ before applying the halfway rule.

Problem: Express $22/7$ to 2 decimal places.
Solution: Long division yields $22 \div 7 = 3.1428\dots$. Inspect 3rd digit ($2 < 5$). The second digit remains unchanged. Thus, $22/7 \approx \mathbf{3.14}$.

Premature Truncation: Stopping long division at $2/3 = 0.66$ causes an error, because the third digit $6 \ge 5$ correctly rounds it to $0.67$.

Orbital Mechanics: NASA uses 15 decimal places of $\pi$ ($3.141592653589793$) to navigate space probes across millions of miles with centimeter accuracy.

Absolute Error: The magnitude of deviation between the exact true value and its approximated value:
$\text{Absolute Error} = |\text{True Value} - \text{Approximated Value}|$

Estimate $47 \times 18$: Round $47 \approx 50$ and $18 \approx 20$. Estimated product $= 50 \times 20 = \mathbf{1,000}$. (Exact $= 846$, providing a rapid sanity check).

Problem: True length of a wire is $7.84\text{ m}$. A student measures it as $7.8\text{ m}$. Find the absolute error.
Solution: $\text{Absolute Error} = |7.84 - 7.80| = \mathbf{0.04\text{ m}}$ (or $4\text{ cm}$).

Negative Errors: Absolute error is always positive ($|\Delta x|$), measuring magnitude of deviation regardless of sign.

Budget Planning: Shoppers estimate grocery bills by rounding item costs to the nearest ten rupees to stay strictly within wallet budgets.