\(% Differentiation % https://tex.stackexchange.com/a/60546/ \newcommand{\Diff}{\mathop{}\!\mathrm{d}} \newcommand{\DiffFrac}[2]{\frac{\Diff #1}{\Diff #2}} \newcommand{\DiffOp}[1]{\frac{\Diff}{\Diff #1}} \newcommand{\Ndiff}[1]{\mathop{}\!\mathrm{d}^{#1}} \newcommand{\NdiffFrac}[3]{\frac{\Ndiff{#1} #2}{\Diff {#3}^{#1}}} \newcommand{\NdiffOp}[2]{\frac{\Ndiff{#1}}{\Diff {#2}^{#1}}} % Evaluation \newcommand{\LEvalAt}[2]{\left.#1\right\vert_{#2}} \newcommand{\SqEvalAt}[2]{\left[#1\right]_{#2}} % Epsilon & Phi \renewcommand{\epsilon}{\varepsilon} \renewcommand{\phi}{\varphi} % Sets \newcommand{\NN}{\mathbb{N}} \newcommand{\ZZ}{\mathbb{Z}} \newcommand{\QQ}{\mathbb{Q}} \newcommand{\RR}{\mathbb{R}} \newcommand{\CC}{\mathbb{C}} \newcommand{\PP}{\mathbb{P}} \renewcommand{\emptyset}{\varnothing} % Probabililty \DeclareMathOperator{\Cov}{Cov} \DeclareMathOperator{\Corr}{Corr} \DeclareMathOperator{\Var}{Var} \DeclareMathOperator{\Expt}{E} \DeclareMathOperator{\Prob}{P} % Distribution \DeclareMathOperator{\Binomial}{B} \DeclareMathOperator{\Poisson}{Po} \DeclareMathOperator{\Normal}{N} \DeclareMathOperator{\Exponential}{Exp} \DeclareMathOperator{\Geometric}{Geo} \DeclareMathOperator{\Uniform}{U} % Complex Numbers \DeclareMathOperator{\im}{Im} \DeclareMathOperator{\re}{Re} % Missing Trigonometric & Hyperbolic functions \DeclareMathOperator{\arccot}{arccot} \DeclareMathOperator{\arcsec}{arcsec} \DeclareMathOperator{\arccsc}{arccsc} \DeclareMathOperator{\sech}{sech} \DeclareMathOperator{\csch}{csch} \DeclareMathOperator{\arsinh}{arsinh} \DeclareMathOperator{\arcosh}{arcosh} \DeclareMathOperator{\artanh}{artanh} \DeclareMathOperator{\arcoth}{arcoth} \DeclareMathOperator{\arsech}{arsech} \DeclareMathOperator{\arcsch}{arcsch} % UK Notation \DeclareMathOperator{\cosec}{cosec} \DeclareMathOperator{\arccosec}{arccosec} \DeclareMathOperator{\cosech}{cosech} \DeclareMathOperator{\arcosech}{arcosech} % Paired Delimiters \DeclarePairedDelimiter{\ceil}{\lceil}{\rceil} \DeclarePairedDelimiter{\floor}{\lfloor}{\rfloor} \DeclarePairedDelimiter{\abs}{\lvert}{\rvert} \DeclarePairedDelimiter{\ang}{\langle}{\rangle} % Vectors \newcommand{\vect}[1]{\mathbf{#1}} \newcommand{\bvect}[1]{\overrightarrow{#1}} % https://tex.stackexchange.com/a/28213 % \DeclareMathSymbol{\ii}{\mathalpha}{letters}{"10} % \DeclareMathSymbol{\jj}{\mathalpha}{letters}{"11} % \newcommand{\ihat}{\vect{\hat{\ii}}} % \newcommand{\jhat}{\vect{\hat{\jj}}} \newcommand{\ihat}{\textbf{\^{ı}}} \newcommand{\jhat}{\textbf{\^{ȷ}}} \newcommand{\khat}{\vect{\hat{k}}} % Other Functions \DeclareMathOperator{\sgn}{sgn} \DeclareMathOperator{\tr}{tr} % Other Math Symbols \DeclareMathOperator{\modulo}{mod} \newcommand{\divides}{\mid} \newcommand{\notdivides}{\nmid} \newcommand{\LHS}{\text{LHS}} \newcommand{\RHS}{\text{RHS}} \newcommand{\degree}{^{\circ}}\)

2018.3.1 Question 1

  1. By differentiation with respect to \(\beta \), we have \[ f'(\beta ) = 1 + \frac {1}{\beta ^2} + \frac {2}{\beta ^3}. \]

    If \(f'(t) = 0\), we must have \[ t^3 + t + 2 = 0. \]

    Therefore, \[ (t + 1)(t^2 - t + 2) = 0, \] and hence the only real root to this is \(t = -1\), since \((-1)^2 - 2 \cdot 4 < 0\).

    This means the only stationary point of \(y = f(\beta )\) is \((-1, f(-1) = -1)\).

    For the limiting behaviour of the function, we first look at the case where \(\beta > 0\). As \(\beta \to \infty \), we have \(f(\beta ) \to \beta \) from below. As \(\beta \to 0^{+}\), we have \(f(\beta ) \to -\frac {1}{\beta } - \frac {1}{\beta ^2} \to -\infty \).

    When \(\beta < 0\), we use the substitution \(t = -\frac {1}{\beta }\) to make the behaviours more convincing, and hence \[ f(\beta ) = \beta + t - t^2. \]

    As \(\beta \to 0^{-}\), we have \(t \to \infty \), and \(f(\beta ) \to t - t^2 \to -\infty \). As \(\beta \to -\infty \), we have \(t \to 0^{+}\), and \(f(\beta ) \to \beta \) from above, since \(t - t^2 = t (1 - t) > 0\) when \(0 < t < 1\).

    This means the curve \(y = f(\beta )\) is as below.

    xyOyy(−== 1,fx−(x1))

    Similarly, by differentiation with respect to \(\beta \), we have \[ g'(\beta ) = 1 - \frac {3}{\beta ^2} + \frac {2}{\beta ^3}. \]

    If \(g'(t) = 0\), we must have \[ t^3 - 3t + 2 = 0. \]

    Therefore, \[ (t - 1)^2(t + 2) = 0, \] and hence the real roots to this is \(t = 1\) and \(t = -2\).

    This means the stationary points of \(y = g(\beta )\) is \((1, g(1) = 3)\) and \((-2, g(-2) = -\frac {15}{4})\).

    For the limiting behaviour of the function, we first look at the case where \(\beta > 0\). We consider the substitution \(t = -\frac {1}{\beta }\) to make the behaviours more convincing, and hence \[ g(\beta ) = \beta - 3t - t^2. \]

    As \(\beta \to \infty \), \(t \to 0^{-}\), and hence \(f(\beta ) \to \beta \) from below, since \(-3t - t^2 = - t(t + 3) > 0\) for \(-3 < t < 0\). As \(\beta \to 0^{+}\), \(t \to -\infty \), and hence \(f(\beta ) \to -3t - t^2 \to -\infty \).

    When \(\beta < 0\), we have as \(\beta \to 0^{-}\), \(f(\beta ) \to -\infty \). As \(\beta \to -\infty \), \(f(\beta ) \to \beta \) from below.

    This means the curve \(y = g(\beta )\) is as below.

    (    15)
xyOyy(1− = =,32gx),(−x)4

  2. By Vieta’s Theorem, we have \(u + v = -\alpha \), and \(uv = \beta \). Hence, \[ u + v + \frac {1}{uv} = -\alpha + \frac {1}{\beta }, \] and \[ \frac {1}{u} + \frac {1}{v} + uv = \frac {u + v}{uv} + uv = - \frac {\alpha }{\beta } + \beta . \]
  3. By the given condition, we have \[ -\alpha + \frac {1}{\beta } = -1 \iff \alpha = 1 + \frac {1}{\beta }. \]

    Hence, \begin {align*} \frac {1}{u} + \frac {1}{v} + uv & = - \frac {\alpha }{\beta } + \beta \\ & = - \frac {1 + \frac {1}{\beta }}{\beta } + \beta \\ & = \frac {\beta ^2 - 1 - \frac {1}{\beta }}{\beta } \\ & = \beta - \frac {1}{\beta } - \frac {1}{\beta ^2} \\ & = f(\beta ). \end {align*}

    Also, since \(u, v\) are both real, we have \begin {align*} \alpha ^2 - 4\beta & = \left (1 + \frac {1}{\beta }\right )^2 - 4\beta \\ & = 1 + \frac {2}{\beta } + \frac {1}{\beta ^2} - 4\beta \\ & = \frac {-4 \beta ^3 + \beta ^2 + 2\beta + 1}{\beta ^2} \\ & \geq 0. \end {align*}

    Multiplying both sides by \(-\beta ^2\) (which flips the sign) gives \begin {align*} 4\beta ^3 - \beta ^2 - 2\beta - 1 & \leq 0 \\ (\beta - 1)(4\beta ^2 + 3\beta + 1) & \leq 0. \end {align*}

    This cubic has exactly one real root \(\beta = 1\), so the solution to this inequality is \(\beta \leq 1\) and \(\beta \neq 0\).

    Notice that \(f\) is increasing on \((0, 1] \subset (0, \infty )\). Therefore, for \(\beta > 0\), \[ f(\beta ) \leq f(1) = 1 - 1 - 1 = -1. \]

    When \(\beta < 0\), we have \[ f(\beta ) \leq f(-1) = -1. \]

    So for the range of \(\beta \) in this question, we always have \(f(\beta ) \leq -1\). But we also have \(\frac {1}{u} + \frac {1}{v} + uv \leq -1\) as shown before. These gives us exactly our desired statement.

  4. By the given condition, we have \[ -\alpha + \frac {1}{\beta } = 3 \iff \alpha = -3 + \frac {1}{\beta }. \]

    Hence, \begin {align*} \frac {1}{u} + \frac {1}{v} + uv & = - \frac {\alpha }{\beta } + \beta \\ & = - \frac {-3 + \frac {1}{\beta }}{\beta } + \beta \\ & = \beta + \frac {3}{\beta } - \frac {1}{\beta ^2} \\ & = g(\beta ). \end {align*}

    Also, since \(u, v\) are both real, we have \(\beta \leq 1\) and \(\beta \neq 0\) as well.

    \(g\) must be increasing on \((0, 1]\). Hence, for \(\beta > 0\), we have \[ g(\beta ) \leq g(1) = 3. \]

    When \(\beta < 0\), we have \[ g(\beta ) \leq g(-2) = -\frac {15}{4}. \]

    Since \(3 > -\frac {15}{4}\), we can conclude that the maximum value of \(\frac {1}{u} + \frac {1}{v} + uv\) is \(3\), and it is taken when \(\beta = 1\), which corresponds to \(\alpha = -2\).