Summary
This lesson studies the decimal writing of rational numbers. It shows that the digits of a quotient arise from the remainders of long division, defines the value of an infinite decimal expansion by means of nested intervals and the Archimedean property, and proves the three theorems that characterize it: an irreducible fraction has a terminating expansion if and only if its denominator has no prime factors other than and ; the expansion of every rational number is terminating or repeating, with a period of length less than the denominator; and, conversely, every repeating expansion represents a rational number, whose generating fraction is obtained from the sum of a geometric progression. From this follow the equality between 0.999… and , the characterization of irrational numbers by their non-repeating expansions and, finally, scientific notation, significant figures and the bound on the error that rounding commits.
Learning objectives
- Obtain the decimal expansion of a fraction by long division and justify, with the remainders, why each digit lies between and and why the expansion approximates the fraction.
- Decide, from the factorization of the denominator, whether a fraction has a terminating or a repeating expansion, and bound the length of its period.
- Find the generating fraction of a repeating expansion and justify rigorously equalities such as that of 0.999… with .
- Write numbers in scientific notation, round them to a given number of significant figures and bound the absolute and relative errors committed.
Long division and decimal digits
In the lesson on the rational numbers fractions and their operations were studied; in the lesson on positional numeral systems, the writing of natural numbers with digits. The decimal expansion brings both together: it extends positional notation to the right of the decimal point, with negative powers of . Thus the number written 0.375 is, by definition, a sum of tenths, hundredths and thousandths:
In general, a terminating decimal expansion consists of an integer part and a finite number of digits , each between and , and represents the following number:
Since the sign of a rational number is written separately, it suffices to study the positive rationals. To find the expansion of a fraction of natural numbers one uses long division, which is nothing other than repeated division with remainder: first is divided by , which gives the integer part and a remainder ; then, at each step, the last remainder is multiplied by and divided by . The quotient of that division is the next digit, and its remainder feeds the step that follows:
We compute the digits of . At each step we write the dividend, which is ten times the previous remainder, the digit obtained and the new remainder. The figure shows the staircase of divisions and, above it, the digits as they accumulate.
Proof
Since , the integer part is and the first remainder is the numerator itself, . We multiply the remainder by and divide by : the tenths digit is .
With the remainder we obtain the hundredths digit.
With the remainder , the thousandths digit.
With the remainder , the fourth digit.
With the remainder , the fifth.
With the remainder , the sixth; and the new remainder is , the same one we started with.
From here on the division repeats step by step: what remains to be divided, multiplied by , is the same quotient shifted six places. The digits form a block that repeats without end.
The expansion of is therefore 0.428571428571…, with period 428571. The example contains the whole lesson in germ: the digits come from the remainders, the remainders are less than the divisor and, as soon as one repeats, everything that follows repeats. The first fact worth securing is that the digits so obtained are genuine digits and that they approximate the fraction as closely as one wishes.
Let and be natural numbers, and let , and, for each natural number , and be the numbers given by the long division of by . Then each is a digit, between and , and for every natural number the following equality holds, in which the last summand is nonnegative and less than .
The proof proceeds by induction on . The figure accompanies the argument with : each new digit chooses one of the ten parts of the previous interval, which is magnified to show it.
Proof
Each is a digit: since the previous remainder is less than , the dividend is less than , and the quotient is less than .
For , with no digits, the equality is the first division with remainder, divided by .
Inductive step: the last summand is rewritten with the next division, , and split into two fractions.
Substituting into the equality of order gives the one of order ; by induction, the equality holds for every .
Since , the last summand is nonnegative and less than : the fraction lies at a distance less than from its truncation of order , the number formed by the integer part and the first digits.
The lemma says that lies in the interval that begins at and has length , with the left endpoint included and the right one excluded; these intervals, which will be studied in detail in the lesson on the real line, are written as follows:
Loading the content…
Completed
Sign in to save your progress