多项式科技

本文主要是介绍多项式科技，对大家解决编程问题具有一定的参考价值，需要的程序猿们随着小编来一起学习吧！

不知道多久才能做完多项式全家桶 qaq

多项式乘法

快速傅里叶变换 (FFT)

直接上链接（

快速傅里叶变换(FFT)详解 - 自为风月马前卒

总的来说就是先 DFT 从系数表示法到点值表示法，再 IDFT 从点值表示法到系数表示法。

简单说一下不太理解的，在 DFT 中 \(\omega_n^k = -\omega_n^{k+\frac{n}{2}}\)，其实就是在单位圆上旋转了 \(180°\)。

感性理解 DFT 和 IDFT 是逆运算，所以 \(\omega_n\) 就是相反数。

Code

#include <bits/stdc++.h>
#define ll long long
#define db double
#define gc getchar
#define pc putchar

using namespace std;

namespace IO
{
    template <typename T>
    void read(T &x)
    {
        x = 0; bool f = 0; char c = gc();
        while(!isdigit(c)) f |= c == '-', c = gc();
        while(isdigit(c)) x = x * 10 + c - '0', c = gc();
        if(f) x = -x;
    }

    template <typename T>
    void write(T x)
    {
        if(x < 0) pc('-'), x = -x;
        if(x > 9) write(x / 10);
        pc('0' + x % 10);
    }
}
using namespace IO;

struct Complex
{
    db x, y;

    Complex(db _x = 0.0, db _y = 0.0)
    {
        x = _x, y = _y;
    }

    Complex operator + (const Complex b) const
    {
        return Complex(x + b.x, y + b.y);
    }

    Complex operator - (const Complex b) const
    {
        return Complex(x - b.x, y - b.y);
    }

    Complex operator * (const Complex b) const
    {
        return Complex(x * b.x - y * b.y, x * b.y + y * b.x);
    }
};

const int N = 1e6 + 5;
const db pi = acos(-1.0);

int n, m;
Complex f[N << 2], g[N << 2];
int len = 1, bit, rev[N << 2];

void FFT(Complex a[], int type)
{
    for(int i = 0; i < len; i++)
        if(i < rev[i]) swap(a[i], a[rev[i]]);
    for(int mid = 1; mid < len; mid <<= 1)
    {
        Complex wn(cos(pi / mid), type * sin(pi / mid));
        for(int i = 0; i < len; i += (mid << 1))
        {
            Complex w(1, 0);
            for(int j = 0; j < mid; j++, w = w * wn)
            {
                Complex x = a[i + j], y = w * a[i + mid + j];
                a[i + j] = x + y;
                a[i + mid + j] = x - y;
            }
        }
    }
    if(type == -1)
        for(int i = 0; i < len; i++)
            a[i].x /= len;
    return;
}

int main()
{
    read(n), read(m);
    for(int i = 0; i <= n; i++) read(f[i].x);
    for(int i = 0; i <= m; i++) read(g[i].x);

    while(len <= n + m) len <<= 1, bit++;
    for(int i = 0; i < len; i++)
        rev[i] = (rev[i >> 1] >> 1) | ((i & 1) << (bit - 1));

    FFT(f, 1);
    FFT(g, 1);
    for(int i = 0; i < len; i++) f[i] = f[i] * g[i];
    FFT(f, -1);

    for(int i = 0; i <= n + m; i++)
        printf("%lld ", (ll)(f[i].x + 0.5));
    pc('\n');

    return 0;
}
// A.S.

快速数论变换 (NTT)

继续上链接（

快速数论变换(NTT)小结 - 自为风月马前卒

就是把 FFT 中的 \(\omega_n\) 改为了 \(g^{\frac{p-1}{n}}\)，其中 \(g\) 为模数的原根。

\[\omega_n \equiv g^{\frac{p-1}{n}}\ (\bmod p) \]

~~证明就算了~~

然后在 IDFT 时就感性理解地取一下逆元就行了（

Code

#include <bits/stdc++.h>
#define ll long long
#define db double
#define gc getchar
#define pc putchar
#define swap(a, b) a ^= b ^= a ^= b

using namespace std;

namespace IO
{
    template <typename T>
    void read(T &x)
    {
        x = 0; bool f = 0; char c = gc();
        while(!isdigit(c)) f |= c == '-', c = gc();
        while(isdigit(c)) x = x * 10 + c - '0', c = gc();
        if(f) x = -x;
    }

    template <typename T>
    void write(T x)
    {
        if(x < 0) pc('-'), x = -x;
        if(x > 9) write(x / 10);
        pc('0' + x % 10);
    }
}
using namespace IO;

const int N = 1e6 + 5;
const int p = 998244353;
const int G = 3;
const int Gi = 332748118;

int n, m, len = 1, bit;
ll f[N << 2], g[N << 2];
int rev[N << 2];

ll qpow(ll a, int b)
{
    ll res = 1;
    while(b)
    {
        if(b & 1) res = res * a % p;
        a = a * a % p, b >>= 1;
    }
    return res % p;
}

void NTT(ll a[], int type)
{
    for(int i = 0; i < len; i++)
        if(i < rev[i]) swap(a[i], a[rev[i]]);
    for(int mid = 1; mid < len; mid <<= 1)
    {
        ll wn = qpow(type == 1 ? G : Gi, (p - 1) / (mid << 1));
        for(int i = 0; i < len; i += (mid << 1))
        {
            ll w = 1;
            for(int j = 0; j < mid; j++, w = w * wn % p)
            {
                ll x = a[i + j], y = w * a[i + mid + j] % p;
                a[i + j] = (x + y) % p;
                a[i + mid + j] = (x - y + p) % p;
            }
        }
    }
    ll inv = qpow(len, p - 2);
    if(type == -1)
        for(int i = 0; i < len; i++)
            f[i] = f[i] * inv % p;
    return;
}

int main()
{
    read(n), read(m);
    for(int i = 0; i <= n; i++) read(f[i]);
    for(int i = 0; i <= m; i++) read(g[i]);

    while(len <= n + m) len <<= 1, bit++;
    for(int i = 0; i < len; i++)
        rev[i] = (rev[i >> 1] >> 1) | ((i & 1) << (bit - 1));

    NTT(f, 1);
    NTT(g, 1);
    for(int i = 0; i < len; i++) f[i] = f[i] * g[i] % p;
    NTT(f, -1);

    for(int i = 0; i <= n + m; i++)
        write(f[i]), pc(' ');
    pc('\n');

    return 0;
}
// A.S.

多项式求逆

对于一个多项式 \(F(x)\) 求一个多项式 \(G(x)\)，满足 \(F(x)*G(x)\equiv 1\ (\bmod x^n)\)，系数对 \(998244353\) 取模。

就是多项式的逆元。

推一下式子

设

\[A(x)*B(x) \equiv 1\ (\bmod x^n)\\ A(x)*C(x)\equiv 1\ (\bmod x^{\frac{n}{2}}) \]

那么

\[A(x)*(B(x)-C(x))\equiv 0\ (\bmod x^{\frac{n}{2}}) \\ B(x)-C(x)\equiv 0\ (\bmod x^{\frac{n}{2}}) \]

我们要把模数改为 \(x^n\)，只需要平方一下

\[[B(x)-C(x)]^2\equiv 0\ (\bmod x^n) \\ B^2(x)-2B(x)*C(x)+C^2(x)\equiv 0\ (\bmod x^n)\\ \]

将等式左边乘上 \(A(x)\) 不影响等号

因为

\[A(x)*B(x)\equiv 1\ (\bmod x^n) \]

所以可以将 \(A(x)*B(x)\) 都去掉

\[B(x)-B(x)*C(x)+A(x)*C^2(x)\equiv 0\ (\bmod x^n) \\ B(x)\equiv B(x)*C(x)-A(x)*C^2(x)\ (\bmod x^n) \]

我们只需要求出 \(C(x)\)，而 \(C(x)\) 与 \(B(x)\) 的形式是一样的，只是模数不一样，所以可以递归求解，当然也可以递推

复杂度 \(O(n\log n)\)

这里写的递推，但是常数好像不是很优秀（

\(bas\) 是当前多项式的项数，\(len\) 是当前多项式乘起来后的项数，也就是 \(2\times bas\)。

Code

#include <bits/stdc++.h>
#define ll long long
#define db double
#define gc getchar
#define pc putchar
#define swap(a, b) a ^= b ^= a ^= b

using namespace std;

namespace IO
{
    template <typename T>
    void read(T &x)
    {
        x = 0; bool f = 0; char c = gc();
        while(!isdigit(c)) f |= c == '-', c = gc();
        while(isdigit(c)) x = x * 10 + c - '0', c = gc();
        if(f) x = -x;
    }

    template <typename T>
    void write(T x)
    {
        if(x < 0) pc('-'), x = -x;
        if(x > 9) write(x / 10);
        pc('0' + x % 10);
    }
}
using namespace IO;

const int N = 1e5 + 5;
const int p = 998244353;
const int G = 3;
const int Gi = 332748118;

int n;
ll a[N << 2], b[2][N << 2];
int rev[N << 2];

ll qpow(ll a, int b)
{
    ll res = 1;
    while(b)
    {
        if(b & 1) res = res * a % p;
        a = a * a % p, b >>= 1;
    }
    return res;
}

void calcrev(int len, int bit)
{
    for(int i = 0; i < len; i++)
        rev[i] = (rev[i >> 1] >> 1) | ((i & 1) << (bit - 1));
}

ll add(ll x, ll y)
{
    return (x + y >= p) ? (x + y - p) : (x + y);
}

ll sub(ll x, ll y)
{
    return add(x, p - y);
}

void NTT(ll a[], int len, int type)
{
    for(int i = 0; i < len; i++)
        if(i < rev[i]) swap(a[i], a[rev[i]]);

    for(int mid = 1; mid < len; mid <<= 1)
    {
        ll wn = qpow(type == 1 ? G : Gi, (p - 1) / (mid << 1));
        for(int i = 0; i < len; i += (mid << 1))
        {
            ll w = 1;
            for(int j = 0; j < mid; j++, w = w * wn % p)
            {
                ll x = a[i + j], y = w * a[i + mid + j] % p;
                a[i + j] = add(x, y);
                a[i + mid + j] = sub(x, y);
            }
        }
    }

    ll inv = qpow(len, p - 2);
    if(type == -1)
        for(int i = 0; i < len; i++)
            a[i] = a[i] * inv % p;
    return;
}

ll X[N << 2], Y[N << 2];

void mul(ll x[], ll y[], int len)
{
    memset(X, 0, sizeof(X)), memset(Y, 0, sizeof(Y));
    for(int i = 0; i < (len >> 1); i++) X[i] = x[i], Y[i] = y[i];
    NTT(X, len, 1);
    NTT(Y, len, 1);
    for(int i = 0; i < len; i++) X[i] = X[i] * Y[i] % p;
    NTT(X, len, -1);
    for(int i = 0; i < len; i++) x[i] = X[i];
    return;
}

void solve()
{
    int k = 0, bas = 1, bit = 1, len = 2;
 
    b[k][0] = qpow(a[0], p - 2);
    while(bas < (n << 1))
    {
        calcrev(len, bit);
        k ^= 1;
        for(int i = 0; i < bas; i++) b[k][i] = add(b[k ^ 1][i], b[k ^ 1][i]);
        mul(b[k ^ 1], b[k ^ 1], len);
        mul(b[k ^ 1], a, len);
        for(int i = 0; i < bas; i++) b[k][i] = sub(b[k][i], b[k ^ 1][i]);
        bas <<= 1, len <<= 1, bit++;
    }

    for(int i = 0; i < n; i++)
        write(b[k][i]), pc(' ');
    pc('\n');

    return;
}

int main()
{
    read(n);
    for(int i = 0; i < n; i++) read(a[i]), a[i] %= p;
    solve();
    return 0;
}
// A.S.

多项式 ln

给定一个多项式 \(A(x)\)，求一个多项式 \(B(x)\)，满足 \(B(x)\equiv \ln A(x)\ (\bmod x^n)\)

设 \(f(x) = \ln x\)

\(\ln\) 不好处理，但是对 \(\ln\) 求导后就很好算了，\(f'(x)=\dfrac{1}{x}\)

所以将同余号两边同时求导，\(B'(x)\equiv f'(A(x))*A'(x)\) （复合函数求导）

因为 \(f'(A(x))=\dfrac{1}{A(x)}\)

所以 \(B'(x)\equiv \dfrac{A'(x)}{A(x)}\)

有了 \(B'(x)\) 后再积分求出 \(B(x)\)

求导公式：\((x^a)'=ax^{a-1}\)

积分公式：\(\int x^adx=\dfrac{1}{a+1}x^{a+1}\)

积分就是求导的逆运算，你会发现把 \(\dfrac{1}{a+1}x^{a+1}\) 求导后就是 \(x^a\)

Code

#include <bits/stdc++.h>
#define ll long long
#define db double
#define gc getchar
#define pc putchar
#define swap(a, b) a ^= b ^= a ^= b

using namespace std;

namespace IO
{
    template <typename T>
    void read(T &x)
    {
        x = 0; bool f = 0; char c = gc();
        while(!isdigit(c)) f |= c == '-', c = gc();
        while(isdigit(c)) x = x * 10 + c - '0', c = gc();
        if(f) x = -x;
    }

    template <typename T>
    void write(T x)
    {
        if(x < 0) pc('-'), x = -x;
        if(x > 9) write(x / 10);
        pc('0' + x % 10);
    }
}
using namespace IO;

const int N = 1e5 + 5;
const int p = 998244353;
const int G = 3;
const int Gi = 332748118;

int n;
ll f[N << 2], g[N << 2];
ll a[N << 2], b[N << 2];

ll qpow(ll a, int b)
{
    ll res = 1;
    while(b)
    {
        if(b & 1) res = res * a % p;
        a = a * a % p, b >>= 1;
    }
    return res;
}

ll add(ll x) {return x >= p ? x - p : x; }
ll sub(ll x) {return x < 0 ? x + p : x; }
void Copy(ll *x, ll *y, int len) {for(int i = 0; i < len; i++) x[i] = y[i]; }

int rev[N << 2];

void calcrev(int len, int bit)
{
    for(int i = 0; i < len; i++)
        rev[i] = (rev[i >> 1] >> 1) | ((i & 1) << (bit - 1));
}

void NTT(ll *a, int len, int type)  //快速数论变换
{
    for(int i = 0; i < len; i++)
        if(i < rev[i]) swap(a[i], a[rev[i]]);

    for(int mid = 1; mid < len; mid <<= 1)
    {
        ll wn = qpow(type == 1 ? G : Gi, (p - 1) / (mid << 1));
        for(int i = 0; i < len; i += (mid << 1))
        {
            ll w = 1;
            for(int j = 0; j < mid; j++, w = w * wn % p)
            {
                ll x = a[i + j], y = w * a[i + mid + j] % p;
                a[i + j] = add(x + y);
                a[i + mid + j] = sub(x - y);
            }
        }
    }

    ll leni = qpow(len, p - 2);
    if(type == -1)
        for(int i = 0; i < len; i++)
            a[i] = a[i] * leni % p;
    return;
}

ll X[N << 2], Y[N << 2];

void mul(ll *x, ll *y, int len) //多项式乘法
{
    memset(X, 0, sizeof(X)), memset(Y, 0, sizeof(Y));
    Copy(X, x, len >> 1), Copy(Y, y, len >> 1);

    NTT(X, len, 1);
    NTT(Y, len, 1);
    for(int i = 0; i < len; i++) X[i] = X[i] * Y[i] % p;
    NTT(X, len, -1);

    Copy(x, X, len);
}

ll inv[2][N << 2];

void Inv(ll *x, ll *y, int n)   //多项式求逆
{
    int bas = 1, len = 2, bit = 1, k = 0;

    inv[k][0] = qpow(x[0], p - 2);
    while(bas < (n << 1))
    {
        calcrev(len, bit);
        k ^= 1;
        for(int i = 0; i < bas; i++) inv[k][i] = add(inv[k ^ 1][i] + inv[k ^ 1][i]);
        mul(inv[k ^ 1], inv[k ^ 1], len);
        mul(inv[k ^ 1], x, len);
        for(int i = 0; i < bas; i++) inv[k][i] = sub(inv[k][i] - inv[k ^ 1][i]);
        bas <<= 1, len <<= 1, bit++;
    }

    Copy(y, inv[k], n);
}

void Differential(ll *x, ll *y, int n)  //求导
{
    for(int i = 1; i < n; i++)
        y[i - 1] = i * x[i] % p;
    y[n - 1] = 0;
}

void Integral(ll *x, ll *y, int n)  //积分
{
    for(int i = 1; i < n; i++)
        y[i] = x[i - 1] * qpow(i, p - 2) % p;
    y[0] = 0;
}

int calclen(int n)
{
    int len = 1;
    while(len <= (n << 1)) len <<= 1;
    return len;
}

void Ln(ll *x, ll *y, int n)
{
    Differential(x, a, n);
    Inv(x, b, n);
    mul(a, b, calclen(n));
    Integral(a, y, n);
}

int main()
{
    read(n);
    for(int i = 0; i < n; i++) read(f[i]);

    Ln(f, g, n);
    
    for(int i = 0; i < n; i++)
        write(g[i]), pc(' ');
    pc('\n');

    return 0;
}
// A.S.

这篇关于多项式科技的文章就介绍到这儿，希望我们推荐的文章对大家有所帮助，也希望大家多多支持为之网！

Java教程