Let's say we're given N jobs and K workers to do those jobs. But for some jobs we need 2 employees, while for some we need just one. Also the employees can't do all jobs. For example worker 1 can do jobs 1,2 and 5, while not jobs 3 and 4. Also if we hire worker 1 to do job 1, then we want him to do jobs 2 and 5, since we've already paid him.
So for example let's say we have 5 jobs and 6 workers. For jobs 1,2 and 4 we need 2 men, while for jobs 3 and 5 we need just one. And here's the list of the jobs every worker can do and the wage he requires.
Worker 1 can do jobs 1,3,5 and he requires 1000 dollars.
Worker 2 can do jobs 1,5 and he requires 2000 dollars.
Worker 3 can do jobs 1,2 and he requires 1500 dollars.
Worker 4 can do jobs 2,4 and he requires 2500 dollars.
Worker 5 can do jobs 4,5 and he requires 1500 dollars.
Worker 6 can do jobs 3,5 and he requires 1000 dollars.
After little calculation and logical thinking we can conclude that we have to hire workers 1,3,4 and 5, which means that the minimum wage we need to pay is: 1000+1500+2500+1500=5500 dollars.
But how we can find an efficient algorithm that will output that amount? This somehow reminds me of the Hungarian Algorithm, but all those additional constrains makes it impossible for me to apply it.
We can represent a state of all jobs as a number in a ternary system(2-two people remaing, 1-one person remaining and 0 if it is already done). Now we can compute f(mask, k) = the smallest cost to hire some workers among the first k in such a way that the state of remaining jobs is mask. Transitions are as follows: we either go to (mask, k + 1)(not hiring the current worker) or we go to (new_mask, k + 1)(in this case we pay this worker his salary and let him do all the jobs he can). The answer is f(0, K).
The time complexity is O(3^N * K * N).
Here is an idea how to optimize it further(and get rid of the N factor). Let's assume that the current mask is mask and the man can do jobs from another mask'. We could actually simply add mask to mask', but there is one problem: the positions where there was 2 in the mask and 1 in mask' will get broken. But we can fix: for each mask, let's precompute a binary mask allowed_mask that contain all position where the digit is not 2. For each man and for each allowed_mask we can precompute that mask' value. Now each transition is just one addition:
for i = 0 ... k - 1
for mask = 0 ... 3^n - 1
allowed_mask = precomputed_allowed_mask[mask]
// make a transition to (i + 1, mask + add_for_allowed_mask[i][allowed_mask])
// make a transition to (i + 1, mask)
Note that there are only 2^n allowed masks. So the time complexity of this solution is O(3^N * N + T * 2^N * K * N + T * 3^N * K)(the first term is for precomputing allowed_masks for all ternary mask, the second one is for precomputing mask' for all allowed_masks and people, and the last is for dp itself).
Related
Thor is playing a game where there are N levels and M types of available weapons. The levels are numbered from 0 to N-1 and the weapons are numbered from 0 to M-1. He can clear these levels in any order. In each level, some subset of these M weapons is required to clear this level. If in a particular level, he needs to buy x new weapons, he will pay x^2 coins for it. Also note that he can carry all the weapons he has currently to the next level. Initially, he has no weapons. Can you find out the minimum coins required such that he can clear all the levels?
Input Format
The first line of input contains 2 space separated integers:
N = the number of levels in the game
M = the number of types of weapons
N lines follow. The ith of these lines contains a binary string of length M. If the jth character of this string is 1, it means we need a weapon of type j to clear the ith level.
Constraints
1 <= N <= 20
1 <= M <= 20
Output Format
Print a single integer which is the answer to the problem.
Sample TestCase 1
Input
1 4
0101
Output
4
Explanation
There is only one level in this game. We need 2 types of weapons - 1 and 3. Since, initially, Thor has no weapons he will have to buy these, which will cost him 2^2 = 4 coins.
Sample TestCase 2
Input
3 3
111
001
010
Output
3
Explanation
There are 3 levels in this game. The 0th level (111) requires all 3 types of weapons. The 1st level (001) requires only weapon of type 2. The 2nd level requires only weapon of type 1. If we clear the levels in the given order (0-1-2), total cost = 3^2 + 0^2 + 0^2 = 9 coins. If we clear the levels in the order 1-2-0, it will cost = 1^2 + 1^2 + 1^2 = 3 coins, which is the optimal way.
The beauty of Gassa's answer is partly in the fact that if a different state can be reached by oring one of the levels' bitstring masks with the current state, we are guaranteed that achieving the current state did not include visiting this level (since otherwise those bits would already be set). This means checking a transition from one state to another by adding a different bitmask, guarantees we are looking at an ordering that did not yet include that mask. So a formulation like Gassa's could work: let f(st) represent the cost of acheiving state st, then:
f(st) = min(
some known cost of f(st),
f(prev_st) + (popcount(prev_st | level) - popcount(prev_st))^2
)
for all level and prev_st that or to st
So I try to solve the following task:
Develop an CREW PRAM algorithm for counting the odd numbers of a sequence of integers x_1,x_2,...x_n.
n is the number of processors - the complexity should be O(log n) and log_2 n is a natural number
My solution so far:
Input: A:={x_1,x_2,...,x_n} Output:=oddCount
begin
1. global_read(A(n),a)
2. if(a mod 2 != 0) then
oddCount += 1
The problem is, due to CREW I am not allowed to use multiple write instructions at the same time oddCount += 1 is reading oddCount and then writes oddCount + 1, so there would be multiple writes.
Do I have to do something like this
Input: A:={x_1,x_2,...,x_n} Output:=oddCount
begin
1. global_read(A(n),a)
2. if(a mod 2 != 0) then
global_write(1, B(n))
3. if(n = A.length - 1) then
for i = 0 to B.length do
oddCount += B(i)
So first each process determines wether it is a odd or even number and the last process calculates the sum? But how would this affect the complexity and is there a better solution?
Thanks to libik I came to this solution: (n starts with 0)
Input: A:={x_1,x_2,...,x_n} Output:=A(0):=number off odd numbers
begin
1. if(A(n) mod 2 != 0) then
A(n) = 1
else
A(n) = 0
2. for i = 1 to log_2(n) do
if (n*(2^i)+2^(i-1) < A.length)
A(n*(2^i)) += A(n*(2^i) + (2^(i-1)))
end
i = 1 --> A(n * 2): 0 2 4 6 8 10 ... A(n*2 + 2^0): 1 3 5 7 ...
i = 2 --> A(n * 4): 0 4 8 12 16 ... A(n*4 + 2^1): 2 6 10 14 18 ...
i = 3 --> A(n * 8): 0 8 16 24 32 ... A(n*8 + 2^2): 4 12 20 28 36 ...
So the first if is the 1st Step and the for is representing log_2(n)-1 steps so over all there are log_2(n) steps. Solution should be in A(0).
Your solution is O(n) as there is for cycle that has to go through all the numbers (which means you dont utilize multiple processors at all)
The CREW means you cannot write into the same cell (in your example cell=processor memory), but you can write into multiple cells at once.
So how to do it as fast as possible?
At initialization all processors start with 1 or 0 (having odd number or not)
In first round just sum the neighbours x_2 with x_1, then x_4 with x_3 etc.
It will be done in O(1) as every second processor "p_x" look to "p_x+1" processor in parallel and add 0 or 1 (is there odd number or not)
Then in processors p1,p3,p5,p7.... you have part of solution. Lets do this again but now with p1 looks to p3, p5 looks to p7 and p_x looks to o_x+2
Then you have part of the solution only in processors p1, p5, p9 etc.
Repeat the process. Every step the number of processors halves, so you need log_2(n) steps.
If this would be real-life example, there is often calculated cost of synchronization. Basically after each step, all processors have to synchronize themselves so they now, they can do the second step (as you run the described code in each processor, but how do you know if you can already add number from processor p_x, because you can do it after p_x finished work).
You need either some kind of "clock" or synchronization.
At this example, the final complexity would be log(n)*k, where k is the complexity of synchronization.
The cost depends on machine, or definition. One way how to notify processors that you have finished is basically the same one as the one described here for counting the odd numbers. Then it would also cost k=log(n) which would result in log^2(n)
I have some problems with defining a algorithm that will calculate a ranking number for a dentist.
Assume, we have three different dentists:
dentist number 1: Got 125 patients and out of the 125 patients the
dentist have booked a time with 75 of them. 60% of them got a time.
dentist number 2: Got 5 patients and out of the 5 patients the
dentist have booked a time with 4 of them. 80% of them got a time.
dentist number 3: Got 25 patients and out of the 14 patients the
dentist have booked a time with 14 of them. 56% got a time.
If we use the formula:
patients booked time with / totalpatients * 100
it will not be the right way to calculate the ranking, as we will get an output of the higher percentage is, the better the dentist is, but it's wrong. By doing it in that way, the dentists would have a ranking:
dentist number 2 would have a ranking of 1. (80% got a time).
dentist number 1 would have a ranking of 2 (60% got a time).
dentist number 3 would have a ranking of 3. (56% got a time).
But, it should be in this way:
dentist number 1 = ranking 1
dentist number 2 = ranking 2
dentist number 3 = ranking 3
I don't know to make a algorithm that also takes the amount of patients as a factor to the ranking-calculation.
It is quite arbitrary how you define what makes a better dentist in terms of number of patients and the percentage of those that have an appointment with them.
Let's call the number of patients P, the number of those that have an appointment A, and the function determining how "good" a dentist is f. So f would be a function of P and A: f(P, A).
One component of f could indeed be what you already calculated: A/P.
Another component would have to be P, but I would think that the effect on f(P, A) of increasing P with 1 would be much higher for a low P, than for a high P, so this component should not be a linear function. It would also be practical if this component would have a value between 0 and 1, just like the other component.
Taking all this together, I suggest this definition of f, which will give a number between 0 and 1:
f(P,A) = 1/3 * P/(10 + P) + 2/3 * A/P
For the different dentists, this results in:
1: 1/3 * 125/135 + 2/3 * 75/125 = 0.7086419753...
2: 1/3 * 5/15 + 2/3 * 4/5 = 0.6444444444...
3: 1/3 * 25/35 + 2/3 * 14/25 = 0.6114285714...
You could play a bit with the constant factors in the formula, like increasing the term 10. Or you could change the factors 1/3 and 2/3 making sure that their sum is 1.
This is just one way to do it. There are an infinity of other ways...
Closed. This question is opinion-based. It is not currently accepting answers.
Want to improve this question? Update the question so it can be answered with facts and citations by editing this post.
Closed 4 years ago.
Improve this question
I was wondering what would be the right answer for Question 2-44 in Skiena's Algorithm Design Manual (2nd ed.)
The question is the following:
We have 1,000 data items to store on 1,000 nodes. Each node can store
copies of exactly three different items. Propose a replication scheme
to minimize data loss as nodes fail. What is the expected number of
data entries that get lost when three random nodes fail?
I was thinking about node n having data item from n, n+1 & n+2.
So if 3 consecutive nodes are lost then we lose 1 item.
Is there a better solution?
The approach you propose is not bad but also take a look here. The ideas used in RAID may give you some ideas. For instance if you have 2 data items, than having storage for 3 items you can recover any of them if the other fails. The idea is quite simple - you store the items in 2 nodes and the xor of their bits in the third item. I believe if you utilize this idea you will be able to have more then 3 backups of a single data item(i.e. more then 3 nodes have to fail in order to loose the information).
I thought of methods like RAID levels but Skiena says "each node can store copies of exactly three different items." Even though XOR'red bit patterns of two separate data can be stored in the same amount of space, I did not think that it was something the problem was looking for.
So, I started with what the OP thought of: Store the three copies of each data to its next two neighbors in a striped fashion. For example, the following is for when N==6 and the data are the integers from 0 to 5 (4 and 5 wrap around and use the nodes 0 and 1):
nodes: 0 1 2 3 4 5
===========
copy 0 -> 0 1 2 3 4 5
copy 1 -> 5 0 1 2 3 4
copy 2 -> 4 5 0 1 2 3
Of all the 20 combinations of three-node failures, there are six that lose exactly one piece of data. For example; when nodes 1, 2, and 3 fail, the data 1 gets lost:
===========
0 X X X 4 5
5 X X X 3 4
4 X X X 2 3
Similar for each other data, making 6 of the 20 combinations lose data. Since Skiena does not describe what "data loss" means for the application: Does the loss of a single data point mean that the entire collection is wasted, or losing a single data point is acceptable and is better than losing two?
If the loss of even a single data point means that the entire collection is wasted, then we can do better. Three times better! :)
Instead of distributing the copies of data to the right-hand nodes in a striped fashion, define groups of three nodes that share data. For example, let 0, 1, and 2 share their data and 3, 4, and 5 share their data:
nodes: 0 1 2 3 4 5
===========
copy 0 -> 0 1 2 3 4 5
copy 1 -> 2 0 1 5 3 4
copy 2 -> 1 2 0 4 5 3
This time, there are only 2 of the 20 combinations produce data loss ever. Data 0, 1, and 2 are lost together when nodes 0, 1, and 2 fail:
===========
x x x 3 4 5
x x x 5 3 4
x x x 4 5 3
And data 3, 4, and 5 are lost together when nodes 3, 4, and 5 fail:
===========
0 1 2 x x x
2 0 1 x x x
1 2 0 x x x
That amounts to just 2 of the 20 combinations of three-node failures. When the same nodes share same data, it effectively merges data losses into fewer number of combinations.
Ali
Let,
D = {1,...,d_i,...,d} denote the data items and d_i a given data element
N = {1,...,n_k,...,n} denote the storage cluster and n_k a given storage node.
We say d_i is stored by n_k, loosely denoted by d_i \in n_k.
My replication model has the following assumptions:
1- Every data item must be stored at least in one given node during initialization. I.e.:
Exist at least one 1 <= k <=n s.t. P(d_i \in n_k) = 1.
2- From (1), at initialization time, the probability of d_i to be in a given node is at least 1/n. I.e.:
For any data item 1 <= i <= d and a random node n, P(d_i \in n) >= 1/n.
Given the problem statement, by design, we want to have this distribution uniform across the data set.
3- Lastly, by design, the probability of a data item d_i to be in a given node n should be independent between data items. I.e.:
P(d_i \in n | d_j \in n) = P(d_i \in n)
This is because we don't assume the probability of node failure is independent between adjacent nodes (e.g.: in datacenters adjacent nodes be sharing the same network switch, etc).
From these assumptions, I proposed the following replication model (for the problem instance where d = n and each node stores exactly 3 distinct data items).
(1) Perform a random permutation of data set.
(2) Using a sliding window of length 3 and stride 1, rotate over the shuffled data set and map the data items to each node.
E.g.:
D = {A,B,C,D}
N = {1,2,3,4}
(1) {C, B, A, D}
(2) 1 -> {C, B, A}, 2 -> {B, A, D}, 3-> {A, D, C}, 4-> {D, C, B}
the random shuffling will ensure independent (3) and uniform distribution (2). While the sliding window of stride 1 guarantees (1).
Let's denote, the sliding window of a given node n_k as the ordered set w_k = {w_k1, w_k2, w_k3}. n_k is said to be the master node for w_k1 (first element of w_k). Any other node n_j containing w_k1 is a replica node. N.B.: the proposed replication model guarantees only one master node for any d_i, while the number of replica nodes depends on the window length.
In the example above: n_1 is the master node for C and n_3 and n_4 replica nodes.
Back to the original problem, given this schema, we can state the probability of data loss is the lost of the master node and all replicas for a given data item.
P(d_i is lost) = P(master node for d_i fails and replica 1 fails and replica 2 fails).
without formal proof, an unbiased random permutation in step (1) above would result
P(d_i is lost) = P(master node for d_i fails) * P(replica 1 fails) * P(replica 2 fails).
again, the random permutation is a heuristic to abstract the joint distribution for nodes failure.
From assumptions (2) and (3), P(d_i is lost) = c, for any d_i, at initialization time.
That said for d = n = 1000 and replication factor of 3 (i.e.: window length equals 3).
P(d_i is lost) = 1/1000 * 1/999 * 1/998 ~ 10^-9
Your approach seems essentially correct but can benefit from a failover strategy. Notice that Prof. Skiena has asked "to minimize data loss as nodes fail" which suggests that failing nodes will be a common occurrence.
You may want to have a look at consistent hashing.
Also, there is a great post by reddit engineers about the perils of not using consistent hashing (instead using a fixed MOD hashing).
I was looking to have members submit their top-10 list of something, or their top 10 rankings, then have some algorithm combine the results. Is there something out there like that?
Thanks!
Ahhhh, that's open-ended alright. Let's consider a simple case where only two people vote:
1 ALPHA
2 BRAVO
3 CHARLIE
1 ALPHA
2 DELTA
3 BRAVO
We can't go purely by count... ALPHA should obviously win, though it has the same votes as BRAVO. Yet, we must avoid a case where just a few first place votes dominate a massive amount of 10th place votes. To do this, I suggest the following:
$score = log($num_of_answers - $rank + 2)
First place would then be worth just a bit over one point, and tenth place would get .3 points. That logarithmic scaling prevents ridiculous dominance, yet still gives weight to rankings. From those example votes (and assuming they were the top 3 of a list of 10), you would get:
ALPHA: 2.08
BRAVO: 1.95
DELTA: .1
CHARLIE: .95
Why? Well, that's subjective. I feel out of a very long list that 4,000 10th place votes is worth more than 1,000 1st place votes. You may scale it differently by changing the base of your log (natural, 2, etc.), or choose a different system.
You could just add up the total for each item of the ranking given by a user and then sort them.
ie:
A = (a,b,c)
B = (a,c,b)
C = (b,a,c)
D = (c,b,a)
E = (a,c,b)
F = (c,a,b)
a = 1 + 1 + 2 + 3 + 1 + 2 = 10
b = 2 + 3 + 1 + 2 + 3 + 3 = 14
c = 3 + 2 + 3 + 1 + 2 + 1 = 12
Thus,
a
c
b
I think you could solve this problem by using a max flow algorithm, to create an aggregate ranking, assuming the following:
Each unique item from the list of items is a node in a graph. E.g. if there are 10 things to vote on, there are 10 nodes.
An edge goes from node *a* to node *b* if *a* is immediately before *b* in a _single user submitted_ ranking.
The last node created from a _single user submitted_ ranking will have an edge pointed at the *sink*
The first node created from a _single user submitted_ ranking will have an incoming edge from the *source*
This should get you an aggregated top-10 list.