As discussed in #464, lazy loading of .app5 files would be a desirable additional feature.
For the SPED datasets, the offsets calculated for use in np.frombuffer should be sufficient for loading data from disk. A similar process could also be employed for STEM datasets.
If this feature was added, it would also make sense to depreciate dryrun, as lazy_loader would do almost the same thing in only a slightly longer compute time.
As discussed in #464, lazy loading of .app5 files would be a desirable additional feature.
For the SPED datasets, the
offsetscalculated for use innp.frombuffershould be sufficient for loading data from disk. A similar process could also be employed for STEM datasets.If this feature was added, it would also make sense to depreciate
dryrun, aslazy_loaderwould do almost the same thing in only a slightly longer compute time.