You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
With the upcoming support for QN indices and fermions in the TTN constructor #122, it would be nice allow for making a random TTN in a given quantum number sector (determined by an input state, passed e.g. as a function).
The text was updated successfully, but these errors were encountered:
b-kloss
changed the title
randomTTN for site-indices with QNs is missing
randomTTN for site-indices with QNs
Jan 14, 2024
b-kloss
changed the title
randomTTN for site-indices with QNsrandomTTN for site-indices with QNs
Jan 14, 2024
I've been wanting to improve the implementation of this for the QN case ever since support for QN conserving QR (and LQ) decomposition was added. It would be good to chat about the algorithm for this & could be ideal to add it to this package then "back port" the implementation to ITensors.jl afterward.
Yeah, definitely something we need. Though it should be called random_ttn to follow the convention that only types (and constructors with the exact name of the type) are capitalized.
Also, a temporary workaround for constructing random TTN with bond dimension larger than one would be to apply a low-depth circuit of random 2-site unitary gates to a product state.
mtfishman
changed the title
randomTTN for site-indices with QNsrandom_ttn for site-indices with QNs
Jan 16, 2024
With the upcoming support for
QN
indices and fermions in theTTN
constructor #122, it would be nice allow for making a randomTTN
in a given quantum number sector (determined by an input state, passed e.g. as a function).The text was updated successfully, but these errors were encountered: