Tuesday, 26 January 2010

Expressing power sum symmetric polynomials in terms of lower degree power sums

Combining the trace formula proposed by Gjergji Zaimi and Qiaochu Yuan,
p_k={rm Tr}begin{pmatrix} e_1 & 1 & cdots & 0 \  -e_2 & 0 & ddots & vdots \  vdots & vdots & ddots & 1 \  (-1)^{N-1}e_N & 0 & cdots & 0 end{pmatrix}^{k},

p_k={rm Tr}begin{pmatrix} e_1 & 1 & cdots & 0 \  -e_2 & 0 & ddots & vdots \  vdots & vdots & ddots & 1 \  (-1)^{N-1}e_N & 0 & cdots & 0 end{pmatrix}^{k},

with the formula quoted by Peter Erskin,
e_n=frac1{n!} begin{vmatrix}p_1 & 1 & 0 & cdots\ p_2 & p_1 & 2 & 0 & cdots \ vdots&& ddots & ddots \ p_{n-1} & p_{n-2} & cdots & p_1 & n-1 \ p_n & p_{n-1} & cdots & p_2 & p_1 end{vmatrix},

Mathematica produces the following expansions of pk:




N=2



p3=frac12p31+frac32p1p2



p4=frac12p41+p21p2+frac12p22



p5=frac14p51+frac54p1p22



p6=frac34p41p2+frac32p21p22+frac14p32



p7=frac18p71frac78p51p2+frac78p31p22+frac78p1p32



p8=frac18p81frac12p61p2frac14p41p22+frac32p21p32+frac18p42



p9=frac116p91frac98p51p22+frac32p31p32+frac916p1p42



p10=frac516p81p2frac54p61p22+frac58p41p32+frac54p21p42+frac116p52



p11=frac132p111+frac1132p91p2frac1116p71p22frac1116p51p32+frac5532p31p42+frac1132p1p52




N=3



p4=frac16p41p21p2+frac12p22+frac43p1p3



p5=frac16p51frac56p31p2+frac56p21p3+frac56p2p3



p6=frac112p61frac14p41p2frac34p21p22+frac14p32+frac13p31p32+p1p2p3+frac13p23



p7=frac136p71frac712p31p22+frac736p41p3+frac712p22p3+frac79p1p23



p8=frac172p81frac118p61p2+frac112p41p22frac12p21p32+frac18p42+frac29p51p3


frac89p31p2p3+frac23p1p22p3+frac89p21p23+frac49p2p23




N=4



p5=frac124p51+frac512p31p2frac58p1p22frac56p21p3+frac56p2p3+frac54p1p4



p6=frac124p61+frac38p41p2frac38p21p22frac18p32frac23p31p3+frac13p23+frac34p21p4+frac34p2p4



p7=frac148p71+frac748p51p2+frac748p31p22frac716p1p22frac724p41p3frac712p21p2p3


+frac724p22p3+frac724p31p4+frac78p1p2p4+frac712p3p4



p8=frac1144p81+frac136p61p2+frac524p41p22frac14p21p22frac116p42


frac19p51p3frac29p31p2p3frac13p1p22p3frac49p21p23+frac49p2p23

+frac112p41p4+frac12p21p2p4+frac14p22p4+frac23p1p3p4+frac14p24.




It seems to me that a nice and compact formula for ak,rho does exist. Indeed,
the coefficients in the above examples are extremely simple.



In particular, I observe that the last terms in each of pk for N=8
have the form
kprodjfrac1jrjrj!prjj,


which corresponds to
ak,rho=kprodjfrac1jrjrj!.

This formula (whose structure resembles the coefficients in the expansion of Schur functions quoted by Peter Erskin) also works for all terms of the type pjpkj at arbitrary N.



Apparently, this is not a general formula, as can be seen from the coefficients in front
of pk1 which do depend on N.
I believe, however, that the general formula for ak,rho with N properly included should not be much more complex than the empirical one above.



Hope this helps.

No comments:

Post a Comment