So this is the classic egg drop problem. I just don't get how the recursion works here. How does it reach the end of the function which returns min+1 every time the recursion takes place?
Note that my very conceptual understanding of recursion may be flawed.
/* Function to get minimum number of trials needed in worst
case with n eggs and k floors */
int eggDrop(int n, int k)
{
// If there are no floors, then no trials needed. OR if there is
// one floor, one trial needed.
if (k == 1 || k == 0)
return k;
// We need k trials for one egg and k floors
if (n == 1)
return k;
int min = INT_MAX, x, res;
// Consider all droppings from 1st floor to kth floor and
// return the minimum of these values plus 1.
for (x = 1; x <= k; x++)
{
res = max(eggDrop(n-1, x-1), eggDrop(n, k-x));
if (res < min)
min = res;
}
return min + 1;
}
There are several base cases for the recursion, when n == 1, when k == 0, and when k == 1.
Each non-base-case call has quite a few recursive calls. So, for example, if we want to call eggDrop(2, 5), we will return something like the value of
min (
// **egg breaks** **egg doesn't break**
max (eggDrop (1, 0), eggDrop (2, 4) ),
max (eggDrop (1, 1), eggDrop (2, 3) ),
max (eggDrop (1, 2), eggDrop (2, 2) ),
max (eggDrop (1, 3), eggDrop (2, 1) ),
max (eggDrop (1, 4), eggDrop (2, 0) ),
)
Note that each of these cases moves us toward one of those base cases. In the first column, we have reduced n by 1, and in the second column, we have reduced k by x,
where x is a positive integer no bigger than k. And since each recursive call moves us closer to the base case, we will eventually bottom out.
(You could probably prove this formally by showing that on each recursive call n + k is strictly smaller than on its parent. I won't bother here; the intuition should be enough.)
That should explain how the recursion actually works.
But note how many recursive calls we're making: k * (k - 1) / 2. That means that this recursive version is not likely to be a very performant solution. Hence this problem is generally solved with dynamic programming techniques.
Related
I want to code this constraint.
d and a in the below code are the subsets of set S with the size of N. For example: (N=5, T=3, S=6), d=[1,2,2,3,1] (the elements of d are the first three digits of S and the size of d is N) and a=[6,4,5,6,4] (the elements of a are the three last digits of set S and the size of a is N).
In the constraint, s should start with d and end with a.
It should be like s[j=1]=1:6, s[j=2]=2:4, s[j=3]=2:5, s[j=4]=3:6, s[j=5]1:4.
I do not know how to deal with this set that depends on the other sets. Can you please help me to code my constraint correctly? The below code is not working correctly.
N = 5
T=3
S=6
Cap=15
Q=rand(1:5,N)
d=[1,2,2,3,1]
a=[6,4,5,6,4]
#variable(model, x[j=1:N,t=1:T,s=1:S], Bin)
#constraint(model, [j= 1:N,t = 1:T, s = d[j]:a[j]], sum(x[j,t,s] * Q[j] for j=1:N) <= Cap)
N, T, S = 5, 3, 6
Q = rand(1:5,N)
d = [1, 2, 2, 3, 1]
a = [6, 4, 5, 6, 4]
using JuMP
model = Model()
#variable(model, x[1:N, 1:T, 1:S], Bin)
#constraint(
model,
[t = 1:T, s = 1:S],
sum(x[j, t, s] * Q[j] for j in 1:N if d[j] <= s < a[j]) <= 15,
)
p.s. There's no need to post multiple comments and questions:
Coding arrays in constraint JuMP
You should also consider posting on the Julia discourse instead: https://discourse.julialang.org/c/domain/opt/13. It's easier to have a conversation there.
Problem Statement: The Fibonacci word sequence of bit strings is defined as:
F(0) = 0, F(1) = 1
F(n − 1) + F(n − 2) if n ≥ 2
For example : F(2) = F(1) + F(0) = 10, F(3) = F(2) + F(1) = 101, etc.
Given a bit pattern p and a number n, how often does p occur in F(n)?
Input:
The first line of each test case contains the integer n (0 ≤ n ≤ 100). The second line contains the bit
pattern p. The pattern p is nonempty and has a length of at most 100 000 characters.
Output:
For each test case, display its case number followed by the number of occurrences of the bit pattern p in
F(n). Occurrences may overlap. The number of occurrences will be less than 2^63.
Sample input: 6 10 Sample output: Case 1: 5
I implemented a divide and conquer algorithm to solve this problem, based on the hints that I found on the internet: We can think of the process of going from F(n-1) to F(n) as a string replacement rule: every '1' becomes '10' and '0' becomes '1'. Here is my code:
#include <string>
#include <iostream>
using namespace std;
#define LL long long int
LL count = 0;
string F[40];
void find(LL n, char ch1,char ch2 ){//Find occurences of eiher "11" / "01" / "10" in F[n]
LL n1 = F[n].length();
for (int i = 0;i+1 <n1;++i){
if (F[n].at(i)==ch1&&F[n].at(i+1)==ch2) ++ count;
}
}
void find(char ch, LL n){
LL n1 = F[n].length();
for (int i = 0;i<n1;++i){
if (F[n].at(i)==ch) ++count;
}
}
void solve(string p, LL n){//Recursion
// cout << p << endl;
LL n1 = p.length();
if (n<=1&&n1>=2) return;//return if string pattern p's size is larger than F(n)
//When p's size is reduced to 2 or 1, it's small enough now that we can search for p directly in F(n)
if (n1<=2){
if (n1 == 2){
if (p=="00") return;//Return since there can't be two subsequent '0' in F(n) for any n
else find(n,p.at(0),p.at(1));
return;
}
if (n1 == 1){
if (p=="1") find('1',n);
else find('0',n);
return;
}
}
string p1, p2;//if the last character in p is 1, we can replace it with either '1' or '0'
//p1 stores the substring ending in '1' and p2 stores the substring ending in '0'
for (LL i = 0;i<n1;++i){//We replace every "10" with 1, "1" with 0.
if (p[i]=='1'){
if (p[i+1]=='0'&&(i+1)!= n1){
if (p[i+2]=='0'&&(i+2)!= n1) return;//Return if there are two subsequent '0'
p1.append("1");//Replace "10" with "1"
++i;
}
else {
p1.append("0");//Replace "1" with "0"
}
}
else {
if (p[i+1]=='0'&&(i+1)!= n1){//Return if there are two subsequent '0'
return;
}
p1.append("1");
}
}
solve(p1,n-1);
if (p[n1-1]=='1'){
p2 = p1;
p2.back() = '1';
solve(p2,n-1);
}
}
main(){
F[0] = "0";F[1] = "1";
for (int i = 2;i<38;++i){
F[i].append(F[i-1]);
F[i].append(F[i-2]);
}//precalculate F(0) to F(37)
LL t = 0;//NumofTestcases
int n; string p;
while (cin >> n >> p) {
count = 0;
solve(p,n);
cout << "Case " << ++t << ": " << count << endl;
}
}
The above program works fine, but with small inputs only. When i submitted the above program to codeforces i got an answer wrong because although i shortened the pattern string p and reduces n to n', the size of F[n'] is still very large (n'>=50). How can i modify my code to make it works in this case, or is there another approach (such as dynamic programming?). Many thanks for any advice.
More details about the problem can be found here: https://codeforces.com/group/Ir5CI6f3FD/contest/273369/problem/B
I don't have time now to try to code this up myself, but I have a suggested approach.
First, I should note, that while that hint you used is certainly accurate, I don't see any straightforward way to solve the problem. Perhaps the correct follow-up to that would be simpler than what I'm suggesting.
My approach:
Find the first two ns such that length(F(n)) >= length(pattern). Calculating these is a simple recursion. The important insight is that every subsequent value will start with one of these two values, and will also end with one of them. (This is true for all adjacent values -- for any m > n, F(m) will begin either with F(n) or with F(n - 1). It's not hard to see why.)
Calculate and cache the number of occurrences of the pattern in this these two Fs, but whatever index shifting technique makes sense.
For F(n+1) (and all subsequent values) calculate by adding together
The count for F(n)
The count for F(n - 1)
The count for those spanning both F(n) and F(n - 1). We can achieve that by testing every breakdown of pattern into (nonempty) prefix and suffix values (i.e., splitting at every internal index) and counting those where F(n) ends in prefix and F(n - 1) starts with suffix. But we don't have to have all of F(n) and F(n - 1) to do this. We just need the tail of F(n) and the head of F(n - 1) of the length of the pattern. So we don't need to calculate all of F(n). We just need to know which of those two initial values our current one ends with. But the start is always the predecessor, and the end oscillates between the previous two. It should be easy to keep track.
The time complexity then should be proportional to the product of n and the length of the pattern.
If I find time tomorrow, I'll see if I can code this up. But it won't be in C -- those years were short and long gone.
Collecting the list of prefix/suffix pairs can be done once ahead of time
Mostly I have heard that if you can make a recursion code , you can convert it to a Dynamic programming code, but what is the need to do the same ? And how to convert a recursion code to DP ?
In dynamic programming there are 2 approaches, top-down and bottom-up.
lets take Fibonacci sequence as an example:
f(0) = 0 : x = 1,
f(1) = 1 : x = 1,
f(x) = f(x-1) + f(x-2) : x > 1
The top-down approach:
It uses recursion + memoization(storing the calculated states to avoid the recalculation):
int memo[1000];//initialized by zeroes
int f(int x) {
if (x == 0 || x == 1) return 1;
if (memo[x] != 0) return memo[x]; //trying to avoid recalculation
memo[x] = f(x - 1) + f(x - 2); //storing the result
return memo[x];
}
As you notice here to calculate the value f(x) we have to break it down into
f(x-1) and f(x-2), this why it is called top-down.
The bottom-up approach:
It uses loops(for,while...) rather than recursion and stores the values inside an array:
int memo[1000];
int bottom_up(int x) {
memo[0] = 1;
memo[1] = 1;
for (int i = 2; i < 1000; i++)
memo[i] = memo[i - 1] + memo[i - 2];
}
As you notice we calculate the values of Fibonacci sequence starting from the smaller values up to the bigger ones and this is why it is called bottom-up.
Converting the code from recursion to loops is considered converting the recursive code to an iterative code.
The recursive code will call itself multiple times and you should know that each function call will be stored inside the stack of your memory, so it is preferred to use the iterative approach as it will be better for memory and performance.
I have spent a lot of time to learn about implementing/visualizing dynamic programming problems using iteration but I find it very hard to understand, I can implement the same using recursion with memoization but it is slow when compared to iteration.
Can someone explain the same by a example of a hard problem or by using some basic concepts. Like the matrix chain multiplication, longest palindromic sub sequence and others. I can understand the recursion process and then memoize the overlapping sub problems for efficiency but I can't understand how to do the same using iteration.
Thanks!
Dynamic programming is all about solving the sub-problems in order to solve the bigger one. The difference between the recursive approach and the iterative approach is that the former is top-down, and the latter is bottom-up. In other words, using recursion, you start from the big problem you are trying to solve and chop it down to a bit smaller sub-problems, on which you repeat the process until you reach the sub-problem so small you can solve. This has an advantage that you only have to solve the sub-problems that are absolutely needed and using memoization to remember the results as you go. The bottom-up approach first solves all the sub-problems, using tabulation to remember the results. If we are not doing extra work of solving the sub-problems that are not needed, this is a better approach.
For a simpler example, let's look at the Fibonacci sequence. Say we'd like to compute F(101). When doing it recursively, we will start with our big problem - F(101). For that, we notice that we need to compute F(99) and F(100). Then, for F(99) we need F(97) and F(98). We continue until we reach the smallest solvable sub-problem, which is F(1), and memoize the results. When doing it iteratively, we start from the smallest sub-problem, F(1) and continue all the way up, keeping the results in a table (so essentially it's just a simple for loop from 1 to 101 in this case).
Let's take a look at the matrix chain multiplication problem, which you requested. We'll start with a naive recursive implementation, then recursive DP, and finally iterative DP. It's going to be implemented in a C/C++ soup, but you should be able to follow along even if you are not very familiar with them.
/* Solve the problem recursively (naive)
p - matrix dimensions
n - size of p
i..j - state (sub-problem): range of parenthesis */
int solve_rn(int p[], int n, int i, int j) {
// A matrix multiplied by itself needs no operations
if (i == j) return 0;
// A minimal solution for this sub-problem, we
// initialize it with the maximal possible value
int min = std::numeric_limits<int>::max();
// Recursively solve all the sub-problems
for (int k = i; k < j; ++k) {
int tmp = solve_rn(p, n, i, k) + solve_rn(p, n, k + 1, j) + p[i - 1] * p[k] * p[j];
if (tmp < min) min = tmp;
}
// Return solution for this sub-problem
return min;
}
To compute the result, we starts with the big problem:
solve_rn(p, n, 1, n - 1)
The key of DP is to remember all the solutions to the sub-problems instead of forgetting them, so we don't need to recompute them. It's trivial to make a few adjustments to the above code in order to achieve that:
/* Solve the problem recursively (DP)
p - matrix dimensions
n - size of p
i..j - state (sub-problem): range of parenthesis */
int solve_r(int p[], int n, int i, int j) {
/* We need to remember the results for state i..j.
This can be done in a matrix, which we call dp,
such that dp[i][j] is the best solution for the
state i..j. We initialize everything to 0 first.
static keyword here is just a C/C++ thing for keeping
the matrix between function calls, you can also either
make it global or pass it as a parameter each time.
MAXN is here too because the array size when doing it like
this has to be a constant in C/C++. I set it to 100 here.
But you can do it some other way if you don't like it. */
static int dp[MAXN][MAXN] = {{0}};
/* A matrix multiplied by itself has 0 operations, so we
can just return 0. Also, if we already computed the result
for this state, just return that. */
if (i == j) return 0;
else if (dp[i][j] != 0) return dp[i][j];
// A minimal solution for this sub-problem, we
// initialize it with the maximal possible value
dp[i][j] = std::numeric_limits<int>::max();
// Recursively solve all the sub-problems
for (int k = i; k < j; ++k) {
int tmp = solve_r(p, n, i, k) + solve_r(p, n, k + 1, j) + p[i - 1] * p[k] * p[j];
if (tmp < dp[i][j]) dp[i][j] = tmp;
}
// Return solution for this sub-problem
return dp[i][j];;
}
We start with the big problem as well:
solve_r(p, n, 1, n - 1)
Iterative solution is only to, well, iterate all the states, instead of starting from the top:
/* Solve the problem iteratively
p - matrix dimensions
n - size of p
We don't need to pass state, because we iterate the states. */
int solve_i(int p[], int n) {
// But we do need our table, just like before
static int dp[MAXN][MAXN];
// Multiplying a matrix by itself needs no operations
for (int i = 1; i < n; ++i)
dp[i][i] = 0;
// L represents the length of the chain. We go from smallest, to
// biggest. Made L capital to distinguish letter l from number 1
for (int L = 2; L < n; ++L) {
// This double loop goes through all the states in the current
// chain length.
for (int i = 1; i <= n - L + 1; ++i) {
int j = i + L - 1;
dp[i][j] = std::numeric_limits<int>::max();
for (int k = i; k <= j - 1; ++k) {
int tmp = dp[i][k] + dp[k+1][j] + p[i-1] * p[k] * p[j];
if (tmp < dp[i][j])
dp[i][j] = tmp;
}
}
}
// Return the result of the biggest problem
return dp[1][n-1];
}
To compute the result, just call it:
solve_i(p, n)
Explanation of the loop counters in the last example:
Let's say we need to optimize the multiplication of 4 matrices: A B C D. We are doing an iterative approach, so we will first compute the chains with the length of two: (A B) C D, A (B C) D, and A B (C D). And then chains of three: (A B C) D, and A (B C D). That is what L, i and j are for.
L represents the chain length, it goes from 2 to n - 1 (n is 4 in this case, so that is 3).
i and j represent the starting and ending position of the chain. In case L = 2, i goes from 1 to 3, and j goes from 2 to 4:
(A B) C D A (B C) D A B (C D)
^ ^ ^ ^ ^ ^
i j i j i j
In case L = 3, i goes from 1 to 2, and j goes from 3 to 4:
(A B C) D A (B C D)
^ ^ ^ ^
i j i j
So generally, i goes from 1 to n - L + 1, and j is i + L - 1.
Now, let's continue with the algorithm assuming that we are at the step where we have (A B C) D. We now need to take into account the sub-problems (which are already calculated): ((A B) C) D and (A (B C)) D. That is what k is for. It goes through all the positions between i and j and computes the sub problems.
I hope I helped.
The problem with recursion is the high number of stack frames that need to be pushed/popped. This can quickly become the bottle-neck.
The Fibonacci Series can be calculated with iterative DP or recursion with memoization. If we calculate F(100) in DP all we need is an array of length 100 e.g. int[100] and that's the guts of our used memory. We calculate all entries of the array pre-filling f[0] and f[1] as they are defined to be 1. and each value just depends on the previous two.
If we use a recursive solution we start at fib(100) and work down. Every method call from 100 down to 0 is pushed onto the stack, AND checked if it's memoized. These operations add up and iteration doesn't suffer from either of these. In iteration (bottom-up) we already know all of the previous answers are valid. The bigger impact is probably the stack frames; and given a larger input you may get a StackOverflowException for what was otherwise trivial with an iterative DP approach.
I have 2 tables of values and want to scale the first one so that it matches the 2nd one as good as possible. Both have the same length. If both are drawn as graphs in a diagram they should be as close to each other as possible. But I do not want quadratic, but simple linear weights.
My problem is, that I have no idea how to actually compute the best scaling factor because of the Abs function.
Some pseudocode:
//given:
float[] table1= ...;
float[] table2= ...;
//wanted:
float factor= ???; // I have no idea how to compute this
float remainingDifference=0;
for(int i=0; i<length; i++)
{
float scaledValue=table1[i] * factor;
//Sum up the differences. I use the Abs function because negative differences are differences too.
remainingDifference += Abs(scaledValue - table2[i]);
}
I want to compute the scaling factor so that the remainingDifference is minimal.
Simple linear weights is hard like you said.
a_n = first sequence
b_n = second sequence
c = scaling factor
Your residual function is (sums are from i=1 to N, the number of points):
SUM( |a_i - c*b_i| )
Taking the derivative with respect to c yields:
d/dc SUM( |a_i - c*b_i| )
= SUM( b_i * (a_i - c*b_i)/|a_i - c*b_i| )
Setting to 0 and solving for c is hard. I don't think there's an analytic way of doing that. You may want to try https://math.stackexchange.com/ to see if they have any bright ideas.
However if you work with quadratic weights, it becomes significantly simpler:
d/dc SUM( (a_i - c*b_i)^2 )
= SUM( 2*(a_i - c*b_i)* -c )
= -2c * SUM( a_i - c*b_i ) = 0
=> SUM(a_i) - c*SUM(b_i) = 0
=> c = SUM(a_i) / SUM(b_i)
I strongly suggest the latter approach if you can.
I would suggest trying some sort of variant on Newton Raphson.
Construct a function Diff(k) that looks at the difference in area between your two graphs between fixed markers A and B.
mathematically I guess it would be integral ( x = A to B ){ f(x) - k * g(x) }dx
anyway realistically you could just subtract the values,
like if you range from X = -10 to 10, and you have a data point for f(i) and g(i) on each integer i in [-10, 10], (ie 21 datapoints )
then you just sum( i = -10 to 10 ){ f(i) - k * g(i) }
basically you would expect this function to look like a parabola -- there will be an optimum k, and deviating slightly from it in either direction will increase the overall area difference
and the bigger the difference, you would expect the bigger the gap
so, this should be a pretty smooth function ( if you have a lot of data points )
so you want to minimise Diff(k)
so you want to find whether derivative ie d/dk Diff(k) = 0
so just do Newton Raphson on this new function D'(k)
kick it off at k=1 and it should zone in on a solution pretty fast
that's probably going to give you an optimal computation time
if you want something simpler, just start with some k1 and k2 that are either side of 0
so say Diff(1.5) = -3 and Diff(2.9) = 7
so then you would pick a k say 3/10 of the way (10 = 7 - -3) between 1.5 and 2.9
and depending on whether that yields a positive or negative value, use it as the new k1 or k2, rinse and repeat
In case anyone stumbles upon this in the future, here is some code (c++)
The trick is to first sort the samples by the scaling factor that would result in the best fit for the 2 samples each. Then start at both ends iterate to the factor that results in the minimum absolute deviation (L1-norm).
Everything except for the sort has a linear run time => Runtime is O(n*log n)
/*
* Find x so that the sum over std::abs(pA[i]-pB[i]*x) from i=0 to (n-1) is minimal
* Then return x
*/
float linearFit(const float* pA, const float* pB, int n)
{
/*
* Algebraic solution is not possible for the general case
* => iterative algorithm
*/
if (n < 0)
throw "linearFit has invalid argument: expected n >= 0";
if (n == 0)
return 0;//If there is nothing to fit, any factor is a perfect fit (sum is always 0)
if (n == 1)
return pA[0] / pB[0];//return x so that pA[0] = pB[0]*x
//If you don't like this , use a std::vector :P
std::unique_ptr<float[]> targetValues_(new float[n]);
std::unique_ptr<int[]> indices_(new int[n]);
//Get proper pointers:
float* targetValues = targetValues_.get();//The value for x that would cause pA[i] = pB[i]*x
int* indices = indices_.get(); //Indices of useful (not nan and not infinity) target values
//The code above guarantees n > 1, so it is safe to get these pointers:
int m = 0;//Number of useful target values
for (int i = 0; i < n; i++)
{
float a = pA[i];
float b = pB[i];
float targetValue = a / b;
targetValues[i] = targetValue;
if (std::isfinite(targetValue))
{
indices[m++] = i;
}
}
if (m <= 0)
return 0;
if (m == 1)
return targetValues[indices[0]];//If there is only one target value, then it has to be the best one.
//sort the indices by target value
std::sort(indices, indices + m, [&](int ia, int ib){
return targetValues[ia] < targetValues[ib];
});
//Start from the extremes and meet at the optimal solution somewhere in the middle:
int l = 0;
int r = m - 1;
// m >= 2 is guaranteed => l > r
float penaltyFactorL = std::abs(pB[indices[l]]);
float penaltyFactorR = std::abs(pB[indices[r]]);
while (l < r)
{
if (l == r - 1 && penaltyFactorL == penaltyFactorR)
{
break;
}
if (penaltyFactorL < penaltyFactorR)
{
l++;
if (l < r)
{
penaltyFactorL += std::abs(pB[indices[l]]);
}
}
else
{
r--;
if (l < r)
{
penaltyFactorR += std::abs(pB[indices[r]]);
}
}
}
//return the best target value
if (l == r)
return targetValues[indices[l]];
else
return (targetValues[indices[l]] + targetValues[indices[r]])*0.5;
}