Examples¶

tinyTT ships with 24 runnable scripts in examples/. All of them are executed by the test suite (tests/test_examples_run.py), so a script that stops working fails CI rather than sitting broken in the repository.

Core¶

Example	What it shows
`basic_usage.py`	Construction, arithmetic, matvec, a solve
`tt_basics.py`	Core TT operations end to end
`tt_helpers.py`	`eye`, `zeros`, `kron`, `cat`, `pad`, `reshape`
`tt_linalg.py`	SVD, QR, rounding, norms
`tt_autograd.py`	Gradient tracking with `watch` / `unwatch`

Solvers¶

Example	What it shows
`tt_solvers.py`	ALS, AMEn, CG, GMRES, BiCGSTAB
`tt_eigensolver.py`	`eig_solve`: ground state against a dense reference, the imaginary-time comparison that motivates it, four eigenpairs by deflation, and a chain past dense reach checked variationally
`tt_dmrg.py`	DMRG for matrix–vector products
`tt_fast_products.py`	Swap-based Hadamard, matvec, matmat

Time evolution¶

Example	What it shows
`mpo_ising_tdvp.py`	Imaginary-time TDVP on an Ising MPO
`heat_equation.py`	Heat equation in QTT with an AMEn solve
`heat_dfi_dfo.py`	3-D heat equation with `step_truncate`, then DFI inertial and DFO gauge momentum
`tt_clora_benchmark.py`	tt-CLoRA: `d`-dimensional heat equation, full-rank against LoRA-compressed cores
`tt_clora_parametric_heat.py`	tt-CLoRA on a parametric 1-D heat equation, dense natural gradient against LoRA factors, with a spectral reference
`plot_clora_params.py`	The accuracy-against-parameters figure from the tt-CLoRA benchmark data

Fitting and regression¶

Example	What it shows
`tt_functional.py`	Scalar and vector-valued `FunctionalTT` regression
`tt_ftt_als.py`	ALS training of a functional TT from data
`tt_vector_valued.py`	A TT-matrix as a trainable linear map
`tt_riemannian_gd.py`	Gradient descent on the fixed-rank TT manifold
`tt_compositional.py`	Compositional TT with a basis function map

QTT and UQ¶

Example	What it shows
`tt_qtt_solve.py`	A linear system solved in QTT format
`tt_qtt_functional.py`	QTT function regression on a tensor grid
`tt_qtt_parametric_darcy.py`	Parametric 2-D Darcy in QTT: Kronecker-sum operator, AMEn solve, sparse FE reference
`tt_uq_adf_darcy.py`	UQ-ADF surrogate for parametric Darcy flow with a KL expansion

Running them¶

cd tinyTT
source venv/bin/activate
PYTHONPATH=. python examples/basic_usage.py

Two of them take arguments: tt_clora_benchmark.py and tt_clora_parametric_heat.py accept --help. The rest run as they are.