Acquisition time stamps
|
acquisition_time_stamp: torch.Tensor = field(default_factory=_float_factory) |
describe the time since midnight. To ensure high enough precision (0.1ms) they are being saved as float64. If they are used in signal models, this can lead to the output to be in double precision althouth most of the time single precision is what users would expect/want.
Suggestion: add warning to signal models if these time stamps are provided as double precision.
Acquisition time stamps
mrpro/src/mrpro/data/AcqInfo.py
Line 165 in 0f42922
describe the time since midnight. To ensure high enough precision (0.1ms) they are being saved as float64. If they are used in signal models, this can lead to the output to be in double precision althouth most of the time single precision is what users would expect/want.
Suggestion: add warning to signal models if these time stamps are provided as double precision.