Add gap detection for `stop_t`, though only report atm
parent
e252f70253
commit
5ab4e5493e
|
@ -29,7 +29,6 @@ from numba import (
|
|||
|
||||
# TODO: for ``numba`` typing..
|
||||
# from ._source import numba_ohlc_dtype
|
||||
from ._sharedmem import ShmArray
|
||||
from ._m4 import ds_m4
|
||||
from .._profile import (
|
||||
Profiler,
|
||||
|
@ -328,7 +327,7 @@ def slice_from_time(
|
|||
|
||||
# compute (presumed) uniform-time-step index offsets
|
||||
i_start_t = round(start_t)
|
||||
read_i_start = (i_start_t - t_first) // step
|
||||
read_i_start = ((i_start_t - t_first) // step) - 1
|
||||
|
||||
i_stop_t = round(stop_t)
|
||||
read_i_stop = (i_stop_t - t_first) // step
|
||||
|
@ -339,20 +338,25 @@ def slice_from_time(
|
|||
# - never allow an end index > the read array len
|
||||
read_i_start = min(
|
||||
max(0, read_i_start),
|
||||
read_i_max,
|
||||
read_i_max - 1,
|
||||
)
|
||||
read_i_stop = max(
|
||||
0,
|
||||
min(read_i_stop, read_i_max),
|
||||
)
|
||||
|
||||
samples = (i_stop_t - i_start_t) // step
|
||||
index_diff = read_i_stop - read_i_start + 1
|
||||
if index_diff > (samples + 3):
|
||||
breakpoint()
|
||||
|
||||
# check for larger-then-latest calculated index for given start
|
||||
# time, in which case we do a binary search for the correct index.
|
||||
# NOTE: this is usually the result of a time series with time gaps
|
||||
# where it is expected that each index step maps to a uniform step
|
||||
# in the time stamp series.
|
||||
i_iv_start = index[read_i_start - 1]
|
||||
t_iv_start = times[read_i_start - 1]
|
||||
i_iv_start = index[read_i_start]
|
||||
t_iv_start = times[read_i_start]
|
||||
if (
|
||||
i_iv_start >= i_first
|
||||
and t_iv_start > i_start_t
|
||||
|
@ -386,7 +390,20 @@ def slice_from_time(
|
|||
# f'diff: {t_diff}\n'
|
||||
# f'REMAPPED START i: {read_i_start} -> {new_read_i_start}\n'
|
||||
# )
|
||||
read_i_start = new_read_i_start
|
||||
read_i_start = new_read_i_start - 1
|
||||
|
||||
t_iv_stop = times[read_i_stop - 1]
|
||||
if (
|
||||
i_stop_t <= t_last
|
||||
and t_iv_stop < i_stop_t
|
||||
):
|
||||
t_diff = stop_t - t_iv_stop
|
||||
print(
|
||||
f"WE'RE CUTTING OUT TIME - STEP:{step}\n"
|
||||
f'calced iv stop:{t_iv_stop} -> stop_t:{stop_t}\n'
|
||||
f'diff: {t_diff}\n'
|
||||
# f'SHOULD REMAP STOP: {read_i_start} -> {new_read_i_start}\n'
|
||||
)
|
||||
|
||||
# read-relative indexes: gives a slice where `shm.array[read_slc]`
|
||||
# will be the data spanning the input time range `start_t` ->
|
||||
|
|
Loading…
Reference in New Issue