def group_reduce_min(val):
"""
First thread of first wave get the result
"""
tid = roc.get_local_id(0)
blksz = roc.get_local_size(0)
wid = tid >> WAVEBITS
lane = tid & (WAVESIZE - 1)
sm_partials = roc.shared.array(WAVESIZE, dtype=dtype)
val = wave_reduce_min(val)
if lane == 0:
sm_partials[wid] = val
roc.barrier()
val = sm_partials[lane] if tid < (blksz //
WAVESIZE) else dtype(POS_INF)
if wid == 0:
val = wave_reduce_min(val)
return val
def udt(output):
global_id = roc.get_global_id(0)
global_size = roc.get_global_size(0)
local_id = roc.get_local_id(0)
group_id = roc.get_group_id(0)
num_groups = roc.get_num_groups(0)
workdim = roc.get_work_dim()
local_size = roc.get_local_size(0)
output[0, group_id, local_id] = global_id
output[1, group_id, local_id] = global_size
output[2, group_id, local_id] = local_id
output[3, group_id, local_id] = local_size
output[4, group_id, local_id] = group_id
output[5, group_id, local_id] = num_groups
output[6, group_id, local_id] = workdim
def kernel_reduce_min(inp, out, nelem):
tid = roc.get_local_id(0)
blkid = roc.get_group_id(0)
blksz = roc.get_local_size(0)
numgroup = roc.get_num_groups(0)
i = blkid * blksz + tid
accum = dtype(POS_INF)
while i < nelem:
accum = min(accum, inp[i])
i += blksz * numgroup
accum = group_reducer(accum)
if tid == 0:
out[blkid] = accum
def roc_uni_kde(support, samples, bandwidth, pdf):
gid = roc.get_group_id(0)
tid = roc.get_local_id(0)
tsz = roc.get_local_size(0)
supp = support[gid]
# all local threads cooperatively computes the energy for a support
energy = 0
for base in range(0, samples.size, tsz):
idx = tid + base
if idx < samples.size:
energy += kernel((samples[idx] - supp) / bandwidth) / bandwidth
# reduce energy
total = group_reduce_sum_float64(energy)
if tid == 0:
pdf[gid] = total / samples.size