學生化全距分佈#

此分佈有兩個形狀參數,\(k>1\)\(\nu>0\),且其支持域為 \(x \geq 0\)

\begin{eqnarray*} f(x; k, \nu) = \frac{k(k-1)\nu^{\nu/2}}{\Gamma(\nu/2)2^{\nu/2-1}} \int_{0}^{\infty} \int_{-\infty}^{\infty} s^{\nu} e^{-\nu s^2/2} \phi(z) \phi(sx + z) [\Phi(sx + z) - \Phi(z)]^{k-2} \,dz \,ds \end{eqnarray*}
\begin{eqnarray*} F(q; k, \nu) = \frac{k\nu^{\nu/2}}{\Gamma(\nu/2)2^{\nu/2-1}} \int_{0}^{\infty} \int_{-\infty}^{\infty} s^{\nu-1} e^{-\nu s^2/2} \phi(z) [\Phi(sq + z) - \Phi(z)]^{k-1} \,dz \,ds \end{eqnarray*}

注意:\(\phi(z)\)\(\Phi(z)\) 分別代表常態 PDF 和常態 CDF。

\(\nu\) 超過 100,000 時,將使用 \(F(x; k, \nu=\infty)\)\(f(x; k, \nu=\infty)\) 的漸近近似

\begin{eqnarray*} F(x; k, \nu=\infty) = k \int_{-\infty}^{\infty} \phi(z) [\Phi(x + z) - \Phi(z)]^{k-1} \,dz \end{eqnarray*}
\begin{eqnarray*} f(x; k, \nu=\infty) = k(k-1) \int_{-\infty}^{\infty} \phi(z)\phi(x + z) [\Phi(x + z) - \Phi(z)]^{k-2} \,dz \end{eqnarray*}

實作:scipy.stats.studentized_range