(
self,
potentials=None,
value=None,
value_linear=None,
value_quad=None,
plan=None,
log=None,
backend=None,
sparse_plan=None,
lazy_plan=None,
status=None,
batch_size=100,
)
| 1092 | """ |
| 1093 | |
| 1094 | def __init__( |
| 1095 | self, |
| 1096 | potentials=None, |
| 1097 | value=None, |
| 1098 | value_linear=None, |
| 1099 | value_quad=None, |
| 1100 | plan=None, |
| 1101 | log=None, |
| 1102 | backend=None, |
| 1103 | sparse_plan=None, |
| 1104 | lazy_plan=None, |
| 1105 | status=None, |
| 1106 | batch_size=100, |
| 1107 | ): |
| 1108 | self._potentials = potentials |
| 1109 | self._value = value |
| 1110 | self._value_linear = value_linear |
| 1111 | self._value_quad = value_quad |
| 1112 | self._plan = plan |
| 1113 | self._log = log |
| 1114 | self._sparse_plan = sparse_plan |
| 1115 | self._lazy_plan = lazy_plan |
| 1116 | self._backend = backend if backend is not None else NumpyBackend() |
| 1117 | self._status = status |
| 1118 | self._batch_size = batch_size |
| 1119 | |
| 1120 | # I assume that other solvers may return directly |
| 1121 | # some primal objects? |
| 1122 | # In the code below, let's define the main quantities |
| 1123 | # that may be of interest to users. |
| 1124 | # An OT solver returns an object that inherits from OTResult |
| 1125 | # (e.g. SinkhornOTResult) and implements the relevant |
| 1126 | # methods (e.g. "plan" and "lazy_plan" but not "sparse_plan", etc.). |
| 1127 | # log is a dictionary containing potential information about the solver |
| 1128 | |
| 1129 | # Dual potentials -------------------------------------------- |
| 1130 |
nothing calls this directly
no test coverage detected