91+991+9991+99991+….
Find the remainder when the greatest integer less than or equal to 10100S is divided by 1000.
解析
Solution 1
Expressing S algebraically, we can write
S=a≥1∑10a−11=a≥1∑b≥1∑10−ab.
The term 10−n will appear exactly τ(n) times in the above sum: this is because the divisor function τ(n) counts the number of ordered pairs (a,b) with a,b≥1 and ab=n. So we have
S=n≥1∑τ(n)10−n,10100S=n≥1∑τ(n)10100−n.
The sum of all terms with n>100 is less than 1, so it is negligible. Hence
After a bit of thinking, you'll realize that for each digit after 0.… is the number of factors of that digit.
So we're basically looking for the number of factors of 98,99,100(and 101 to check for carried ones).
Factors of 98=2⋅72,99=11⋅32,100=22⋅52,101=101,98→2⋅3=6,99→2⋅3=6,100→3⋅3=9,101→2.
So, we get 669 and 2. The two doesn't carry a one, therefore the answer is 669.
~ BenjaminDong01
Solution 3 (Decimals)
We can represent each of the terms as a recurring decimal.
For example, 91=0.1 and 991=0.01. Given this it is trivial to see that the only terms that will contribute to the nth decimal place will be 10a−11 where a is a factor of n.
Now consider multiplying S by 10100 and taking it modulo 1000. This would leave the digits in the 98th, 99th, and 100th places.
So, we find that 98 has 6 factors, 99 has 6 factors, and 100 has 9 factors. 101 is prime, so we don't need to worry about any carrying affecting the 100th decimal place.
Therefore, the answer is 669.
~mathmaniak
Solution 4 (Really thorough)
It is very well known that 10n−1a=0. n-1 zeros00...00a.
Therefore, we can say that
Tn=10n−11=0. n-1 zeros00...001
.
Let Sk=∑n=0k(Tn),
The problem asks us for
⌊10100S∞⌋(mod1000)
.
let Dn be the n−th place value of S∞, before converting. For example, D48=10. Observe that each Tn is periodic with period n. Then, it sbould be obvious that each value of Dkn (k∈Z+) has a 1 contributed from Tn, as Tn "deposits" as 1 every n digits.
We can therefore say that Dn=σ0(n) (σ0(n) is the number of divisors of n. In general, σk(n) is the sum of the k−th powers of the divisors n)
I think we can agree that S∞=∑n=0∞(10nDn)=∑n=0∞(10nσ0(n)) from the definition of place values.
We can also argue that the terms after D100 are inconsequential. This is basically because it is easy enough to see that σ0(n) is much less than 10n−100 if n>100, and summing all of those values is logically not going to be over 1. IE, those terms will never "escape the decimal place". Then, the floor part will just remove that.
So we can say that
⌊10100S∞⌋=10100S100=10100n=0∑100(10nσ0(n))
We want:
10100n=0∑100(10nσ0(n))(mod1000)
Since we are working mod 1000, all of the terms with n≤97 will end up as σ0(n)∗10(r≥3). So, all of the terms just disappear before n=98. We are then left with:
σ0(98)∗10(100−98)+σ0(99)∗10(100−99)+σ0(100)∗10(100−100)=6∗100+6∗10+9=669
Because 98=2∗49=21∗72∴σ0(98)=2∗3=6 and 99=9∗11=32∗112∴σ0(99)=2∗3=6 and 100=4∗25=22∗52∴σ0(100)=3∗3=9
~Stereotypicalmathnerd
Solution 5 (Intuition for Solution 1)
Like in solution 1, we see that we just have the sum ∑k=1∞10k−11. We factor out 10k at the bottom to obtain ∑k=1∞10k(1−10−k)1. We replace 10k1 as 10−k and the result is ∑k=1∞1−10−k10−k. However, that result is just an infinite geometric series! We have that r=10−k and 0<∣r∣<1. The first term of this series is r itself. Thus, the series is rl for some values l, and we can conclude the sum as ∑l=1∞(10−k)l. We determine the value for k using an exterior nested sum to obtain the double sum found in solution 1. Continue as followed.