Generate portable md5 hashes from (arbitrarily nested) dictionaries. These dictionaries can contain arbitrary Python and NumPy data types. The goal of the module is to provide a hash function that can be safely used across different platforms. Its main use is to generate unique identifiers for parameter dictionaries used in parameter scans of neural network simulations.
It exposes a single function to the user
It exposes a single function to the user:
The user can set two global parameters:
To ensure consistency between different systems, the library multiplies floats with the
FLOAT_FACTORand then converts them to integers.
If the float is smaller than the inverse of the
FLOAT_FACTOR, it cannot be safely converted. If
FLOOR_SMALL_FLOATSis set to True, the library will round the float to zero. If set to False, it will throw an error in this case.