Masteries · Jason GlowneyArithmoí Book 1

Masteries

1.2.5 The Dihedral Cone Differential Group ∈3ℂ\in 3\mathbb{C}

We move on to study the de­riv­a­tives for the di­he­dral group D3𝕔D_{3\mathbb{c}} ∈ℂ3×3\in \mathbb{C}^{3 \times 3}, where there are twenty-four el­e­ments in the set A𝕔={Xj,k 𝕔,Yj,k𝕔}:( (j={1,2,3}k={0,1,2,3}))A^{\mathbb{c}} = \left\{ X_{j,k\ }^{\mathbb{c}}{,Y}_{j,k}^{\mathbb{c}} \right\}:\left( \ \binom{j = \left\{ 1,2,3 \right\}}{k = \left\{ 0,1,2,3 \right\}} \right)

Xj,k 𝕔={(e(iπ4+iπk2)0000e(iπ4+iπk2)0e(iπ4+iπk2)0),   (00e(iπ4+iπk2)0e(iπ4+iπk2)0e(iπ4+iπk2)00), (0e(iπ4+iπk2)0e(iπ4+iπk2)0000e(iπ4+iπk2))},   ( (j={1,2,3}k={0,1,2,3})).\begin{array}{r} X_{j,k\ }^{\mathbb{c}} = \left\{ \begin{array}{r} \begin{pmatrix} e^{\left( \frac{i\pi}{4} + \frac{i\pi k}{2} \right)} & 0 & 0 \\ 0 & 0 & e^{\left( \frac{i\pi}{4} + \frac{i\pi k}{2} \right)} \\ 0 & e^{\left( \frac{i\pi}{4} + \frac{i\pi k}{2} \right)} & 0 \end{pmatrix},\ \ \ \begin{pmatrix} 0 & 0 & e^{\left( \frac{i\pi}{4} + \frac{i\pi k}{2} \right)} \\ 0 & e^{\left( \frac{i\pi}{4} + \frac{i\pi k}{2} \right)} & 0 \\ e^{\left( \frac{i\pi}{4} + \frac{i\pi k}{2} \right)} & 0 & 0 \end{pmatrix},\ \\ \begin{pmatrix} 0 & e^{\left( \frac{i\pi}{4} + \frac{i\pi k}{2} \right)} & 0 \\ e^{\left( \frac{i\pi}{4} + \frac{i\pi k}{2} \right)} & 0 & 0 \\ 0 & 0 & e^{\left( \frac{i\pi}{4} + \frac{i\pi k}{2} \right)} \end{pmatrix} \end{array} \right\},\ \ \ \left( \ \binom{j = \left\{ 1,2,3 \right\}}{k = \left\{ 0,1,2,3 \right\}} \right). \end{array}
(1.2.77)
Yj,k𝕔={(e(iπk2)000e(iπk2)000e(iπk2)),   (0e(iπk2)000e(iπk2)e(iπk2)00), (00e(iπk2)e(iπk2)000e(iπk2)0)},   ( (j={1,2,3}k={0,1,2,3}))\begin{array}{r} Y_{j,k}^{\mathbb{c}} = \left\{ \begin{array}{r} \begin{pmatrix} e^{\left( \frac{i\pi k}{2} \right)} & 0 & 0 \\ 0 & e^{\left( \frac{i\pi k}{2} \right)} & 0 \\ 0 & 0 & e^{\left( \frac{i\pi k}{2} \right)} \end{pmatrix},\ \ \ \begin{pmatrix} 0 & e^{\left( \frac{i\pi k}{2} \right)} & 0 \\ 0 & 0 & e^{\left( \frac{i\pi k}{2} \right)} \\ e^{\left( \frac{i\pi k}{2} \right)} & 0 & 0 \end{pmatrix},\ \\ \begin{pmatrix} 0 & 0 & e^{\left( \frac{i\pi k}{2} \right)} \\ e^{\left( \frac{i\pi k}{2} \right)} & 0 & 0 \\ 0 & e^{\left( \frac{i\pi k}{2} \right)} & 0 \end{pmatrix} \end{array} \right\},\ \ \ \left( \ \binom{j = \left\{ 1,2,3 \right\}}{k = \left\{ 0,1,2,3 \right\}} \right) \end{array}
(1.2.78)

We keep in mind that the values of kk can be con­sid­ered as cyclic, mean­ing k mod 4,k\ \text{mod}\ 4,and it should be noted that as the kk index in­creas­es from zero, there is a ro­ta­tion by iπk2\frac{i\pi k}{2} ra­di­ans that is oc­cur­ring. We start with the ex­po­nen­tial map­ping of the first Lie al­ge­bra x1X1,0𝕔x_{1}X_{1,0}^{\mathbb{c}}

G(x1X1,0𝕔) =(e(x12+ix12)000cosh⁡(x12+ix12)sinh⁡(x12+ix12)0sinh⁡(x12+ix12)cosh⁡(x12+ix12)),\begin{array}{r} G\left( {x_{1}X}_{1,0}^{\mathbb{c}} \right)\ = \begin{pmatrix} e^{\left( \frac{x_{1}}{\sqrt{2}} + \frac{ix_{1}}{\sqrt{2}} \right)} & 0 & 0 \\ 0 & \cosh\left( \frac{x_{1}}{\sqrt{2}} + \frac{ix_{1}}{\sqrt{2}} \right) & \sinh\left( \frac{x_{1}}{\sqrt{2}} + \frac{ix_{1}}{\sqrt{2}} \right) \\ 0 & \sinh\left( \frac{x_{1}}{\sqrt{2}} + \frac{ix_{1}}{\sqrt{2}} \right) & \cosh\left( \frac{x_{1}}{\sqrt{2}} + \frac{ix_{1}}{\sqrt{2}} \right) \end{pmatrix}, \end{array}
(1.2.79)

which has the fol­low­ing dif­fer­en­tial form

G′(x1X1,0𝕔)=(−1)14×(e(x12+ix12)000sinh⁡(x12+ix12)cosh⁡(x12+ix12)0cosh⁡(x12+ix12)sinh⁡(x12+ix12)).\begin{array}{r} G^{'}\left( {x_{1}X}_{1,0}^{\mathbb{c}} \right) = ( - 1)^{\frac{1}{4}} \times \begin{pmatrix} e^{\left( \frac{x_{1}}{\sqrt{2}} + \frac{ix_{1}}{\sqrt{2}} \right)} & 0 & 0 \\ 0 & \sinh\left( \frac{x_{1}}{\sqrt{2}} + \frac{ix_{1}}{\sqrt{2}} \right) & \cosh\left( \frac{x_{1}}{\sqrt{2}} + \frac{ix_{1}}{\sqrt{2}} \right) \\ 0 & \cosh\left( \frac{x_{1}}{\sqrt{2}} + \frac{ix_{1}}{\sqrt{2}} \right) & \sinh\left( \frac{x_{1}}{\sqrt{2}} + \frac{ix_{1}}{\sqrt{2}} \right) \end{pmatrix}. \end{array}
(1.2.80)

We see that the de­riv­a­tive be­haves like the (xjXjℝ)′\left( x_{j}X_{j}^{\mathbb{R}} \right)^{'} de­riv­a­tive al­ge­bras, where there is the same switch­ing of the hy­per­bol­ic trigono­met­ric func­tions for the non-axis en­tries, dif­fer­ing given that its com­plex-valued ar­gu­ment leads to a scalar mul­ti­pli­ca­tion of the group by (−1)14( - 1)^{\frac{1}{4}}. G′(x1X1,0𝕔)G^{'}\left( {x_{1}X}_{1,0}^{\mathbb{c}} \right) can also be rep­re­sent­ed as the ex­po­nen­tial map­ping of its de­riv­a­tive al­ge­bra (x1X1,0𝕔)′\left( {x_{1}X}_{1,0}^{\mathbb{c}} \right)^{'}, which is the fol­low­ing phase-shift of the x1X1,0𝕔{x_{1}X}_{1,0}^{\mathbb{c}} al­ge­bra

(x1X1,0𝕔)′=V(ln⁡((eD1,0𝕔)′))(V)−1=(ex1(iπ4)+iπ4000−iπ4ex1(iπ4)+iπ20ex1(iπ4)+iπ2−iπ4),\begin{array}{r} \left( {x_{1}X}_{1,0}^{\mathbb{c}} \right)^{'} = V\left( \ln\left( \left( e^{D_{1,0}^{\mathbb{c}}} \right)^{'} \right) \right)(V)^{- 1} = \begin{pmatrix} e^{x1\left( \frac{i\pi}{4} \right)} + \frac{i\pi}{4} & 0 & 0 \\ 0 & - \frac{i\pi}{4} & e^{x1\left( \frac{i\pi}{4} \right)} + \frac{i\pi}{2} \\ 0 & e^{x1\left( \frac{i\pi}{4} \right)} + \frac{i\pi}{2} & - \frac{i\pi}{4} \end{pmatrix}, \end{array}
(1.2.81)

where the axis entry is phase-shift­ed by iπ4\frac{i\pi}{4} ra­di­ans, the non-axis di­ag­o­nal en­tries by −iπ4- \frac{i\pi}{4} ra­di­ans, and the non-axis vari­able en­tries by iπ2\frac{i\pi}{2} ra­di­ans. Sim­i­lar­ly, the gen­er­al­iza­tion for the (x1X1,k𝕔)′\left( {x_{1}X}_{1,k}^{\mathbb{c}} \right)^{'} de­riv­a­tive al­ge­bras can be rep­re­sent­ed as

(x1X1,k𝕔)′=(ex1(iπ4)+iπk2000−iπ4+iπk2ex1(iπ4)+iπ20ex1(iπ4)+iπ2−iπ4+iπk2),\begin{array}{r} \left( {x_{1}X}_{1,k}^{\mathbb{c}} \right)^{'} = \begin{pmatrix} e^{x_{1}\left( \frac{i\pi}{4} \right)} + \frac{i\pi k}{2} & 0 & 0 \\ 0 & - \frac{i\pi}{4} + \frac{i\pi k}{2} & e^{x_{1}\left( \frac{i\pi}{4} \right)} + \frac{i\pi}{2} \\ 0 & e^{x_{1}\left( \frac{i\pi}{4} \right)} + \frac{i\pi}{2} & - \frac{i\pi}{4} + \frac{i\pi k}{2} \end{pmatrix}, \end{array}
(1.2.82)

which leads to the fol­low­ing gen­er­al­iza­tion for the (xjXj,k𝕔)′\left( {x_{j}X}_{j,k}^{\mathbb{c}} \right)^{'} group of de­riv­a­tive al­ge­bras

(xjXj,k𝕔)′={(ex1(iπ4)+iπk2000−iπ4+iπk2ex1(iπ4)+iπ20ex1(iπ4)+iπ2−iπ4+iπk2),(−iπ4+iπk20ex2(iπ4)+iπ20ex2(iπ4)+iπk20ex2(iπ4)+iπ20−iπ4+iπk2) (−iπ4+iπk2ex3(iπ4)+iπ20ex3(iπ4)+iπ2−iπ4+iπk2000ex3(iπ4)+iπk2)}.\begin{array}{r} \left( {x_{j}X}_{j,k}^{\mathbb{c}} \right)^{'} = \left\{ \begin{array}{r} \begin{pmatrix} e^{x_{1}\left( \frac{i\pi}{4} \right)} + \frac{i\pi k}{2} & 0 & 0 \\ 0 & - \frac{i\pi}{4} + \frac{i\pi k}{2} & e^{x_{1}\left( \frac{i\pi}{4} \right)} + \frac{i\pi}{2} \\ 0 & e^{x_{1}\left( \frac{i\pi}{4} \right)} + \frac{i\pi}{2} & - \frac{i\pi}{4} + \frac{i\pi k}{2} \end{pmatrix},\begin{pmatrix} - \frac{i\pi}{4} + \frac{i\pi k}{2} & 0 & e^{x_{2}\left( \frac{i\pi}{4} \right)} + \frac{i\pi}{2} \\ 0 & e^{x_{2}\left( \frac{i\pi}{4} \right)} + \frac{i\pi k}{2} & 0 \\ e^{x_{2}\left( \frac{i\pi}{4} \right)} + \frac{i\pi}{2} & 0 & - \frac{i\pi}{4} + \frac{i\pi k}{2} \end{pmatrix} \\ \ \begin{pmatrix} - \frac{i\pi}{4} + \frac{i\pi k}{2} & e^{x_{3}\left( \frac{i\pi}{4} \right)} + \frac{i\pi}{2} & 0 \\ e^{x_{3}\left( \frac{i\pi}{4} \right)} + \frac{i\pi}{2} & - \frac{i\pi}{4} + \frac{i\pi k}{2} & 0 \\ 0 & 0 & e^{x_{3}\left( \frac{i\pi}{4} \right)} + \frac{i\pi k}{2} \end{pmatrix} \end{array} \right\}. \end{array}
(1.2.83)

It is also pos­si­ble to find a gen­er­al­iza­tion for the (yjYj,k𝕔)′\left( y_{j}Y_{j,k}^{\mathbb{c}} \right)^{'} de­riv­a­tive al­ge­bras. To show this, we take the (y2Y2,k𝕔)′\left( y_{2}Y_{2,k}^{\mathbb{c}} \right)^{'} ex­am­ple and ex­press it in the di­ag­o­nal­ized form y2αkY2=VDV−1y_{2}\alpha_{k}Y_{2} = VDV^{- 1}

y2αkY2=(1e(−2iπ3)e(2iπ3)1e(2iπ3)e(−2iπ3)111)(y2αke0000y2αke(−2iπ3)000y2αke(2iπ3))(131313e(2iπ3)3e(−2iπ3)313e(−2iπ3)3e(2iπ3)313),\begin{array}{r} y_{2}\alpha_{k}Y_{2} = \begin{pmatrix} 1 & e^{\left( - \frac{2i\pi}{3} \right)} & e^{\left( \frac{2i\pi}{3} \right)} \\ 1 & e^{\left( \frac{2i\pi}{3} \right)} & e^{\left( - \frac{2i\pi}{3} \right)} \\ 1 & 1 & 1 \end{pmatrix}\begin{pmatrix} y_{2}\alpha_{k}e^{0} & 0 & 0 \\ 0 & y_{2}\alpha_{k}e^{\left( - \frac{2i\pi}{3} \right)} & 0 \\ 0 & 0 & y_{2}\alpha_{k}e^{\left( \frac{2i\pi}{3} \right)} \end{pmatrix}\begin{pmatrix} \frac{1}{3} & \frac{1}{3} & \frac{1}{3} \\ \frac{e^{\left( \frac{2i\pi}{3} \right)}}{3} & \frac{e^{\left( - \frac{2i\pi}{3} \right)}}{3} & \frac{1}{3} \\ \frac{e^{\left( - \frac{2i\pi}{3} \right)}}{3} & \frac{e^{\left( \frac{2i\pi}{3} \right)}}{3} & \frac{1}{3} \end{pmatrix}, \end{array}
(1.2.84)

for the vari­ables αk∈ℂ\alpha_{k}\in \mathbb{C}, and y2∈ℝ+y_{2} \in \mathbb{R}^{+}. As we did in the pre­vi­ous sec­tion, we eval­u­ate op­er­a­tions on the di­ag­o­nal­ized matrix of eigen­val­ues DD in the fol­low­ing order: the ex­po­nen­tial map­ping, its de­riv­a­tive, and then the log­a­rithm of the re­sult­ing di­ag­o­nal matrix

D→eD→ddy2[eD]→ln⁡(ddy2[eD]).\begin{array}{r} D \rightarrow e^{D} \rightarrow \frac{d}{dy_{2}}\left\lbrack e^{D} \right\rbrack \rightarrow \ln{\left( \frac{d}{dy_{2}}\left\lbrack e^{D} \right\rbrack \right).} \end{array}
(1.2.85)

After which, the de­com­po­si­tion be­comes

(y2αkY2)′=V(ln⁡(αkey2αk)000ln⁡(αke(−2iπ3)e(y2αke(−2iπ3)))000ln⁡(αke(2iπ3)e(y2αke(2iπ3))))V−1,\begin{array}{r} \left( y_{2}\alpha_{k}Y_{2} \right)^{'} = V\begin{pmatrix} \ln\left( \alpha_{k}e^{y_{2}\alpha_{k}} \right) & 0 & 0 \\ 0 & \ln\left( \alpha_{k}e^{\left( - \frac{2i\pi}{3} \right)}e^{\left( y_{2}\alpha_{k}e^{\left( - \frac{2i\pi}{3} \right)} \right)} \right) & 0 \\ 0 & 0 & \ln\left( \alpha_{k}e^{\left( \frac{2i\pi}{3} \right)}e^{\left( y_{2}\alpha_{k}e^{\left( \frac{2i\pi}{3} \right)} \right)} \right) \end{pmatrix}V^{- 1}, \end{array}
(1.2.86)

where­by sub­sti­tu­tion with αk=e(iπk2)\alpha_{k} = e^{\left( \frac{i\pi k}{2} \right)}, the di­ag­o­nal­ized matrix equals

(ln⁡(e(iπk2)ey2(eiπk2))000ln⁡(e(iπk2)e(−2iπ3)e(y2e(−2iπ3+iπk2)))000ln⁡(e(iπk2)e(2iπ3)e(y2e(2iπ3+iπk2)))),\begin{array}{r} \begin{pmatrix} \ln\left( e^{\left( \frac{i\pi k}{2} \right)}e^{y_{2}\left( \frac{e^{i\pi k}}{2} \right)} \right) & 0 & 0 \\ 0 & \ln\left( e^{\left( \frac{i\pi k}{2} \right)}e^{\left( - \frac{2i\pi}{3} \right)}e^{\left( y_{2}e^{\left( - \frac{2i\pi}{3} + \frac{i\pi k}{2} \right)} \right)} \right) & 0 \\ 0 & 0 & \ln\left( e^{\left( \frac{i\pi k}{2} \right)}e^{\left( \frac{2i\pi}{3} \right)}e^{\left( y_{2}e^{\left( \frac{2i\pi}{3} + \frac{i\pi k}{2} \right)} \right)} \right) \end{pmatrix}, \end{array}
(1.2.87)

and by as­sum­ing the prin­ci­pal-valued log­a­rithm func­tion, it be­comes

(ln⁡(e(iπk2))+y2e(iπk2)000(y2e(−2iπ3+iπk2)−2iπ3+ln⁡(e(iπk2)))000(y2e(2iπ3+iπk2)+2iπ3+ln⁡(e(iπk2)))).\begin{array}{r} \begin{pmatrix} \ln\left( e^{\left( \frac{i\pi k}{2} \right)} \right) + y_{2}e^{\left( \frac{i\pi k}{2} \right)} & 0 & 0 \\ 0 & \left( \begin{array}{r} y_{2}e^{\left( - \frac{2i\pi}{3} + \frac{i\pi k}{2} \right)} \\ - \frac{2i\pi}{3} + \ln\left( e^{\left( \frac{i\pi k}{2} \right)} \right) \end{array} \right) & 0 \\ 0 & 0 & \left( \begin{array}{r} y_{2}e^{\left( \frac{2i\pi}{3} + \frac{i\pi k}{2} \right)} \\ + \frac{2i\pi}{3} + \ln\left( e^{\left( \frac{i\pi k}{2} \right)} \right) \end{array} \right) \end{pmatrix}. \end{array}
(1.2.88)

For VV, the eigen­vec­tor matrix

V=(1e(4iπ3)e(2iπ3)1e(2iπ3)e(4iπ3)111),\begin{array}{r} V = \begin{pmatrix} 1 & e^{\left( \frac{4i\pi}{3} \right)} & e^{\left( \frac{2i\pi}{3} \right)} \\ 1 & e^{\left( \frac{2i\pi}{3} \right)} & e^{\left( \frac{4i\pi}{3} \right)} \\ 1 & 1 & 1 \end{pmatrix}, \end{array}
(1.2.89)

the prod­uct

(y2Y2,k𝕔)′=V(ln⁡(e(iπk2))+y2e(iπk2)000(y2e(−2iπ3+iπk2)−2iπ3+ln⁡(e(iπk2)))000(y2e(−2iπ3+iπk2)−2iπ3+ln⁡(e(iπk2))))V−1,\begin{array}{r} \left( {y_{2}Y}_{2,k}^{\mathbb{c}} \right)^{'} = V\begin{pmatrix} \ln\left( e^{\left( \frac{i\pi k}{2} \right)} \right) + y_{2}e^{\left( \frac{i\pi k}{2} \right)} & 0 & 0 \\ 0 & \left( \begin{array}{r} y_{2}e^{\left( - \frac{2i\pi}{3} + \frac{i\pi k}{2} \right)} \\ - \frac{2i\pi}{3} + \ln\left( e^{\left( \frac{i\pi k}{2} \right)} \right) \end{array} \right) & 0 \\ 0 & 0 & \left( \begin{array}{r} y_{2}e^{\left( - \frac{2i\pi}{3} + \frac{i\pi k}{2} \right)} \\ - \frac{2i\pi}{3} + \ln\left( e^{\left( \frac{i\pi k}{2} \right)} \right) \end{array} \right) \end{pmatrix}V^{- 1}, \end{array}
(1.2.90)

yields the gen­er­al­ized de­riv­a­tive al­ge­bra, (y2Y2,k𝕔)′\left( {y_{2}Y}_{2,k}^{\mathbb{c}} \right)^{'} rep­re­sen­ta­tion

(y2Y2,k𝕔)′=(ln⁡(e(iπk2))y2e(iπk2)+2π33−2π3(e(2iπ3) 3)−2π3(e(2iπ3) 3)ln⁡(e(iπk2))y2e(iπk2)+2π33y2e(iπk2)+2π33−2π3(e(2iπ3) 3)ln⁡(e(iπk2))).\begin{array}{r} \left( {y_{2}Y}_{2,k}^{\mathbb{c}} \right)^{'} = \begin{pmatrix} \ln\left( e^{\left( \frac{i\pi k}{2} \right)} \right) & y_{2}e^{\left( \frac{i\pi k}{2} \right)} + \frac{2\pi}{3\sqrt{3}} & - \frac{2\pi}{3}\left( \frac{e^{\left( \frac{2i\pi}{3} \right)\ }}{\sqrt{3}} \right) \\ - \frac{2\pi}{3}\left( \frac{e^{\left( \frac{2i\pi}{3} \right)\ }}{\sqrt{3}} \right) & \ln\left( e^{\left( \frac{i\pi k}{2} \right)} \right) & y_{2}e^{\left( \frac{i\pi k}{2} \right)} + \frac{2\pi}{3\sqrt{3}} \\ y_{2}e^{\left( \frac{i\pi k}{2} \right)} + \frac{2\pi}{3\sqrt{3}} & - \frac{2\pi}{3}\left( \frac{e^{\left( \frac{2i\pi}{3} \right)\ }}{\sqrt{3}} \right) & \ln\left( e^{\left( \frac{i\pi k}{2} \right)} \right) \end{pmatrix}. \end{array}
(1.2.91)

The gen­er­al­ized form above dif­fers from the dif­fer­en­tial al­ge­bra (y2Y2ℝ)′\left( {y_{2}Y}_{2}^{\mathbb{R}} \right)^{'} by the ad­di­tion of the ln⁡(e(iπk2))\ln\left( e^{\left( \frac{i\pi k}{2} \right)} \right) func­tions for its di­ag­o­nal el­e­ments, and each y2y_{2} vari­able is now mul­ti­plied by e(iπk2).e^{\left( \frac{i\pi k}{2} \right)}.