# ttclass/pack.m

## Syntax

    ttout=pack(tt)


## Description

This subroutine packs all trains from the addition buffer into a single tensor train, but does not perform the recompression. Normally you should not call it directly, use ttclass/shrink.m instead.

## Arguments

    tt  -  tensor train object with unprocessed additions


## Outputs

    ttout - tensor train with additions buffer absorbed
into the cores of the tensor, but not recompressed


## Notes

Additions are expensive in the TT format, and so the tensor train module delays them as much as it can, and only processes them in batches. This function is a part of the processing path.