/
diff_operator_pullback_vector_valued.py
381 lines (307 loc) · 12.5 KB
/
diff_operator_pullback_vector_valued.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
# -*- coding: utf-8 -*-
'''
cf.
[Bö] S.Böcherer, Über die Fourier-Jacobi-Entwickling Siegelscher Eisensteinreihen II,
Mathematische Zeichtschrift, 189 (1985), 81 - 110.
[BSY] S. Böcherer, T. Satoh, T. Yamazaki, On the pullback of a differential operator
and its application to vector valued Eisenstein series,
Comment. Math. Univ. St. Pauli 42 (1992) 1 - 22.
[DIK] N. Dummigan, T. Ibukiyama, H. Katsurada, some Siegel modular standard L-values,
and Shafarevich-Tate groups.
'''
from itertools import combinations
from degree2.standard_l_scalar_valued import (first_elt_of_kern_of_vandermonde,
tpl_to_half_int_mat)
from sage.all import binomial as _binomial
from sage.all import (QQ, ZZ, PolynomialRing, cached_function, factorial,
identity_matrix, matrix, mul, vector, zeta)
from .siegel_series.pullback_of_siegel_eisen import r_n_m_iter
from .siegel_series.siegel_eisenstein import SiegelEisensteinSeries as sess
def binomial(x, m):
return ZZ(_binomial(x, m))
def _permutations_increasing(n, r):
'''
A generator of permutations of n of length r which are non-decreasing.
'''
return combinations(range(n), r)
@cached_function
def permutations_increasing(n, r):
r'''index set of \wedge^r(V)
'''
return list(_permutations_increasing(n, r))
def _concat(a, b):
return tuple(sorted(list(a) + list(b)))
def sub_mat(A, a, b):
return matrix([[A[i, j] for i in a] for j in b])
def _sign(a1, a2):
return mul(mul(-1 for b in a1 if b < a) for a in a2)
@cached_function
def _index_dct(n, r):
return {a: i for i, a in enumerate(permutations_increasing(n, r))}
def sqcap_mul(A, B, n, p, q):
'''
Let A and B be square matrices of size
binomial(n, p) and binomial(n, q).
Return sqcap multiplication defined in [Bö] as a square matrix of size
binomial(n, p + q).
'''
# if p or q is zero, return immediately.
if p == 0:
return B
elif q == 0:
return A
p_dct = _index_dct(n, p)
q_dct = _index_dct(n, q)
p_q_dct = _index_dct(n, p + q)
p_q_lst = permutations_increasing(n, p + q)
res = matrix([[A.base_ring()(0) for _ in p_q_lst] for _ in p_q_lst])
for ad in _permutations_increasing(n, p):
for bd in _permutations_increasing(n, p):
for add in _permutations_increasing(n, q):
for bdd in _permutations_increasing(n, q):
if all(a not in add for a in ad) and all(b not in bdd for b in bd):
a = _concat(ad, add)
b = _concat(bd, bdd)
s = (_sign(ad, add) * _sign(bd, bdd) *
A[p_dct[bd], p_dct[ad]] *
B[q_dct[bdd], q_dct[add]])
res[p_q_dct[b], p_q_dct[a]] += s
return binomial(p + q, p) ** (-1) * res
def bracket_power(A, p):
if p == 0:
return identity_matrix(A.base_ring(), 1)
l = permutations_increasing(A.ncols(), p)
return matrix([[sub_mat(A, a, b).det() for b in l] for a in l])
def _ad_bracket_coeffs(A, a, b):
N = range(A.ncols())
_na = tuple([x for x in N if x not in a])
_nb = tuple([x for x in N if x not in b])
return sub_mat(A, _na, _nb).det() * _sign(a, _na) * _sign(b, _nb)
def ad_bracket(A, p):
if p == A.ncols():
return identity_matrix(A.base_ring(), 1)
l = permutations_increasing(A.ncols(), p)
return matrix([[_ad_bracket_coeffs(A, b, a) for b in l] for a in l])
# matrix([[z11, z12], [z21, z22]]) is (2 pi i)Z_{2} in [Bö].
_Z_ring = PolynomialRing(QQ, names='z11, z12, z21, z22')
# matrix([[dz11, dz12], [dz21, dz22]]) is 2 / (2 pi i)partial_2 in [Bö].
_dZ_ring = PolynomialRing(QQ, names='dz11, dz12, dz21, dz22')
_Z_dZ_ring = PolynomialRing(_Z_ring, names='dz11, dz12, dz21, dz22')
@cached_function
def _z_u_ring_zgens():
return [_Z_U_ring(a) for a in _Z_ring.gens()]
def _from_z_dz_ring_to_diff_op(pol):
pol = _Z_dZ_ring(pol)
d = {tuple(t): _Z_ring(v) for t, v in pol.dict().iteritems()}
return DiffZOperatorElement(d)
def _diff_z_exp(t, pol, r_ls, base_ring=None):
'''Let a, b, c, d = t.
Return
(d/dz11)^a (d/dz12)^b (d/dz21)^c (d/dz22)^d (pol exp(2pi R^t Z))
* exp(- 2pi R^t Z).
Here Z = matrix([[z11, z12], [z21, z22]]) and R = matrix(2, r_ls).
'''
for z, r, a in zip((base_ring(_z) for _z in _Z_ring.gens()), r_ls, t):
pol = base_ring(sum(binomial(a, i) * pol.derivative(z, i) * r ** (a - i)
for i in range(a + 1)))
return pol
class DiffZOperatorElement(object):
'''A polynomial of Z and d/dZ, where Z is a 2 by 2 matrix.
'''
def __init__(self, pol_idcs):
'''pol_idcs is a list of tuples (pol, idx) or a dict whose key is pol
and the value is idx.
Here pol is a polynomial, which is an element of _Z_ring.
idx is a tuple of 4 integers (a, b, c, d).
Each tuple corresponds pol (d/dz11)^a (d/dz12)^b (d/dz21)^c (d/dz22)^d.
'''
if isinstance(pol_idcs, list):
self._pol_idc_dct = {k: v for k, v in pol_idcs if v != 0}
elif isinstance(pol_idcs, dict):
self._pol_idc_dct = {
k: v for k, v in pol_idcs.items() if v != 0}
@property
def pol_idc_dct(self):
return self._pol_idc_dct
def diff(self, pol, r_ls):
'''pol is a polynomial in _Z_ring and R is a 2 by 2 marix.
Return (the derivative of pol * exp(2pi R^t Z)) / exp(R^t Z) as a polynomial.
R = matrix(2, r_ls)
'''
try:
pol = _Z_ring(pol)
except TypeError:
raise NotImplementedError
return sum(v * _diff_z_exp(k, pol, r_ls, base_ring=_Z_ring)
for k, v in self.pol_idc_dct.items())
Z = matrix(2, [_Z_dZ_ring(__a) for __a in _Z_ring.gens()])
dZ = matrix(2, [_Z_dZ_ring(__a) for __a in _dZ_ring.gens()])
def delta_r_q(q, A=None, D=None):
'''
Return delta(r, q) in [DIK], pp. 1312.
'''
n = 2
p = n - q
res = sqcap_mul(bracket_power(A * Z, q) *
bracket_power(
D - ZZ(1) / ZZ(4) * dZ.transpose() * A ** (-1) * dZ, q),
bracket_power(dZ, p), n, q, p)
return res[0, 0] * ZZ(2) ** (-p)
def _C(p, s):
return mul(s + ZZ(i) / ZZ(2) for i in range(p))
def D_tilde(alpha, **kwds):
'''
(2pi)**(-2) * D_tilde(alpha) in [DIK], pp 1312 as an instance of DiffZOperatorElement.
'''
alpha = ZZ(alpha)
res = sum(binomial(2, q) * _C(q, -alpha + 1) ** (-1) * delta_r_q(q, **kwds)
for q in range(3))
return _from_z_dz_ring_to_diff_op(res)
def D_tilde_nu(alpha, nu, pol, r_ls, **kwds):
'''
(2pi)**(-2 nu) * D_tilde_{alpha}^nu(pol * exp(2pi R^t Z)) / exp(- 2pi R^t Z),
where pol is pol polynomial of Z and R = matrix(2, r_ls).
'''
for i in range(nu):
pol = D_tilde(alpha + i, **kwds).diff(pol, r_ls)
return pol
# The repressentation space of Gl2 is homogenous polynomial of u1 and u2.
_U_ring = PolynomialRing(QQ, names='u1, u2')
_Z_U_ring = PolynomialRing(QQ, names='u1, u2, z11, z12, z21, z22')
def _D_D_up_D_down(u1, u2, v1, v2, r_ls, pol):
'''D - D_up - D_down
'''
r11, r12, r21, r22 = [_diff_z_exp(t, pol, r_ls, base_ring=_Z_U_ring) for t in
[(1, 0, 0, 0), (0, 1, 0, 0), (0, 0, 1, 0), (0, 0, 0, 1)]]
return u1 * v1 * r11 + u1 * v2 * r12 + u2 * v1 * r21 + u2 * v2 * r22
def L_operator(k, m, _A, _D, r_ls, pol, us, d_up_down_mlt):
'''
Return (k)_m * Fourier coefficient of
L_tilde^{k, m}(pol exp(2pi block_matrix([[A, R/2], [R^t/2, D]])Z))/
exp(-2pi block_matrix([[A, R/2], [R^t/2, D]])Z).
as an element of _Z_ring or _Z_U_ring.
'''
if m == 0:
return pol
zero = _Z_U_ring(0)
res = zero
u1, u2, u3, u4 = us
for n in range(m // 2 + 1):
pol_tmp = _Z_U_ring(pol)
for _ in range(m - 2 * n):
pol_tmp = _D_D_up_D_down(u1, u2, u3, u4, r_ls, pol_tmp)
for _ in range(n):
pol_tmp *= d_up_down_mlt
pol_tmp *= QQ(factorial(n) * factorial(m - 2 * n) *
mul(2 - k - m + i for i in range(n))) ** (-1)
res += pol_tmp
return res
def _zeta(s):
return zeta(ZZ(s))
def fc_of_pullback_of_diff_eisen(l, k, m, A, D, u3, u4, verbose=False):
'''Return the Fourier coefficient of exp(2pi A Z1 + 2pi D Z2)
of pullback of vector valued Eisenstein series F_{l, (k, m)} in [DIK], pp 1313.
'''
dct = {"A": A, "D": D}
res = _U_ring(0)
es = sess(weight=l, degree=4)
us = list(_U_ring.gens()) + [u3, u4]
# D_up is multiplication by d_up_mlt on p(z2)e(A Z1 + R^t Z12 + D Z2)
v_up = vector(_U_ring, us[:2])
d_up_mlt = v_up * A * v_up
v_down = vector(us[2:])
d_down_mlt = v_down * D * v_down
d_up_down_mlt = d_up_mlt * d_down_mlt
_u1, _u2 = (_Z_U_ring(a) for a in ["u1", "u2"])
for R, mat in r_n_m_iter(A, D):
r_ls = R.list()
pol = D_tilde_nu(l, k - l, QQ(1), r_ls, **dct)
# L_operator is a differential operator whose order <= m,
# we truncate it.
pol = _Z_ring(
{t: v for t, v in pol.dict().iteritems() if sum(list(t)) <= m})
_l_op_tmp = L_operator(k, m, A, D, r_ls, pol *
es.fourier_coefficient(mat), us, d_up_down_mlt)
_l_op_tmp = _U_ring({(m - a, a): _l_op_tmp[_u1 ** (m - a) * _u2 ** a]
for a in range(m + 1)})
res += _l_op_tmp
res = res * QQ(mul(k + i for i in range(m))) ** (-1)
res = res * _zeta(1 - l) * _zeta(1 - 2 * l + 2) * _zeta(1 - 2 * l + 4)
if verbose:
print "Done computation of Fourier coefficient of pullback."
return res
def _pullback_vector(l, D, u3, u4, space_of_cuspforms, verbose=False):
'''Return a vector corresponding to pullback of Eisenstein series.
'''
k = space_of_cuspforms.wt
j = space_of_cuspforms.sym_wt
tpls = space_of_cuspforms.linearly_indep_tuples()
u1, u2 = _U_ring.gens()
if j > 0:
pull_back_fc_dct = {
t: fc_of_pullback_of_diff_eisen(
l, k, j, tpl_to_half_int_mat(t), D, u3, u4, verbose=verbose)
for t in set(t for t, i in tpls)}
pull_back_dct = {(t, i): pull_back_fc_dct[t][u1 ** (j - i) * u2 ** i]
for t, i in tpls}
else:
pull_back_dct = {t: fc_of_pullback_of_diff_eisen(
l, k, j, tpl_to_half_int_mat(t), D, u3, u4,
verbose=verbose) for t in tpls}
pull_back_dct = {k: v.constant_coefficient()
for k, v in pull_back_dct.iteritems()}
return space_of_cuspforms._to_vector(pull_back_dct)
def _u3_u4_gen():
s = 1
while True:
for a in range(s + 1):
yield (a, s - a)
s += 1
def _u3_u4_nonzero(f, t0):
'''
Return (u3, u4, f[t0](u3, u4)) such that f[t0](u3, u4) != 0.
'''
if f.sym_wt > 0:
f_t0_pol = f[t0]._to_pol()
f_t0_pol_val = 0
for u3, u4 in _u3_u4_gen():
if f_t0_pol_val == 0:
x, y = f_t0_pol.parent().gens()
f_t0_pol_val = f_t0_pol.subs({x: u3, y: u4})
u3_val = u3
u4_val = u4
else:
break
else:
f_t0_pol_val = f[t0]
u3_val = u4_val = QQ(1)
return (u3_val, u4_val, f_t0_pol_val)
def algebraic_part_of_standard_l(f, l, space_of_cuspforms, verbose=False):
r'''f: (vector valued) cuspidal eigenform of degree 2 of weight det^k Sym(j).
l: positive even integer such that 2 le l < k - 2.
space_of_cuspforms: space of cusp form that f belongs to.
Return the algebriac part of the standard L of f at l
cf. Katsurada, Takemori Congruence primes of the Kim-Ramakrishnan-Shahidi lift. Theorem 4.1.
'''
k = f.wt
j = f.sym_wt
t0 = f._none_zero_tpl()
D = tpl_to_half_int_mat(t0)
if not (l % 2 == 0 and 2 <= l < k - 2):
raise ValueError
u3_val, u4_val, f_t0_pol_val = _u3_u4_nonzero(f, t0)
pull_back_vec = _pullback_vector(
l + ZZ(2), D, u3_val, u4_val, space_of_cuspforms, verbose=verbose)
T2 = space_of_cuspforms.hecke_matrix(2)
d = space_of_cuspforms.dimension()
vecs = [(T2 ** i) * pull_back_vec for i in range(d)]
ei = [sum(f[t0] * a for f, a in zip(space_of_cuspforms.basis(), v))
for v in vecs]
if j > 0:
ei = [a._to_pol() for a in ei]
chply = T2.charpoly()
nume = first_elt_of_kern_of_vandermonde(chply, f.hecke_eigenvalue(2), ei)
denom = f[t0] * f_t0_pol_val
if j > 0:
denom = denom._to_pol()
return f.base_ring(nume / denom)