xerus tagshttps://git.hemio.de/xerus/xerus/-/tags2016-06-23T16:58:24+02:00https://git.hemio.de/xerus/xerus/-/tags/v4.0.1_condav4.0.1_condaPhilipp Trunschkehttps://git.hemio.de/xerus/xerus/-/tags/v3.0.1v3.0.1Ben Huberben.huber@hemio.dehttps://git.hemio.de/xerus/xerus/-/tags/v3.0.0v3.0.0Ben Huberben.huber@hemio.dehttps://git.hemio.de/xerus/xerus/-/tags/v2.4.0v2.4.0<ul data-sourcepos="1:1-7:78" dir="auto">
<li data-sourcepos="1:1-1:19">2016-06-23 v2.4.0</li>
<li data-sourcepos="2:2-2:110">Introduced nomeclature 'mode'. Marked all functions that will be renamed / removed in v3.0.0 as deprecated.</li>
<li data-sourcepos="3:2-3:122">Added an experimental python wrapper when 'OTHER += -DXERUS_EXPERIMENTAL_PYTHON_WRAPPER' is defined in the config file.</li>
<li data-sourcepos="4:2-4:97">Added TT::degrees_of_freedom function to determine the dimension of the current tangent space.</li>
<li data-sourcepos="5:2-5:65">! Corrected a typo in the class names of the measurement sets.</li>
<li data-sourcepos="6:2-6:92">Added constants xerus::VERSION_MAJOR, VERSION_MINOR, VERSION_REVISION and VERSION_COMMIT.</li>
<li data-sourcepos="7:2-7:78">Improved compilation time and more forgiving default settings for warnings.</li>
</ul>2016-06-23T16:58:24+02:00Benjamin Huberben.huber@hemio.dehttps://git.hemio.de/xerus/xerus/-/tags/v2.3.0v2.3.0<ul data-sourcepos="1:2-7:54" dir="auto">
<li data-sourcepos="1:2-1:113">Added save_to_file and load_from_file functions to store and restore Tensors and TensorNetworks in/from files.</li>
<li data-sourcepos="2:2-2:63">Added SVD based calculate_pseudo_inverse function to Tensor.</li>
<li data-sourcepos="3:2-3:63">Added implementation for perform_trace() for sparse Tensors.</li>
<li data-sourcepos="4:2-4:98">Fixed the pipe operator for xerus::misc::generic_exception to not destroy the type information.</li>
<li data-sourcepos="5:2-5:62">Bug fix in the handling of fixed indices in TensorNetworks.</li>
<li data-sourcepos="6:2-6:76">Several static member function now warn if their return type is not used.</li>
<li data-sourcepos="7:2-7:54">Initial support for compilation with the intel ICC.</li>
</ul>2016-04-04T17:00:17+02:00Benjamin Huberben.huber@hemio.dehttps://git.hemio.de/xerus/xerus/-/tags/v2.2.1v2.2.1<ul data-sourcepos="1:2-1:35" dir="auto">
<li data-sourcepos="1:2-1:35">Added support for 32bit systems.</li>
</ul>2016-03-11T15:40:52+01:00Benjamin Huberben.huber@hemio.dehttps://git.hemio.de/xerus/xerus/-/tags/v2.2.0v2.2.0<ul data-sourcepos="1:2-7:53" dir="auto">
<li data-sourcepos="1:2-1:50">Added Sparse solve and sparse QR factorization.</li>
<li data-sourcepos="2:2-2:61">! Xerus now depends on the SuiteSparseQR (-lspqr) library.</li>
<li data-sourcepos="3:2-3:48">TTNetwork sum now supports sparse components.</li>
<li data-sourcepos="4:2-4:84">TTNetwork entrywise_product now supports sparse components.</li>
<li data-sourcepos="5:2-5:107">! entrywise_product (for TTNetworks) now a free function instead of static member.</li>
<li data-sourcepos="6:2-6:83">TTNetwork entrywise_square now supports sparse components.</li>
<li data-sourcepos="7:2-7:53">Bugfix in Dense/Sparse conversion factor handling.</li>
</ul>2016-03-11T15:40:40+01:00Sebastian Wolfhttps://git.hemio.de/xerus/xerus/-/tags/v2.1.1v2.1.1<ul data-sourcepos="1:2-3:41" dir="auto">
<li data-sourcepos="1:2-1:53">Basic support for fixed indices in TensorNetworks.</li>
<li data-sourcepos="2:2-2:53">Readded TTNetwork constructor from DimensionTuple.</li>
<li data-sourcepos="3:2-3:41">Bugfixes for TTNetwork dyadic_product.</li>
</ul>2016-02-10T11:33:13+01:00Sebastian Wolfhttps://git.hemio.de/xerus/xerus/-/tags/v2.1.0v2.1.0<ul data-sourcepos="1:2-4:111" dir="auto">
<li data-sourcepos="1:2-1:99">! fix_slate() now works correctly for TTTensors and (correctly) throws an error for TTOperators.</li>
<li data-sourcepos="2:2-2:66">Several bug fixes in the handling of TTNetworks of degree zero.</li>
<li data-sourcepos="3:2-3:75">Implemented the Tensor CQ function, which is now used where appropriate.</li>
<li data-sourcepos="4:2-4:111">Added several unit tests to ensure consitent results for dense/sparse Tensors, TensorNetwork and TTNetworks.</li>
</ul>2016-02-05T03:59:17+01:00Sebastian Wolfhttps://git.hemio.de/xerus/xerus/-/tags/v2.0.0v2.0.0<ul data-sourcepos="1:2-12:28" dir="auto">
<li data-sourcepos="1:2-1:67">! Merged FullTensor and SparseTensor into a single Tensor class.</li>
<li data-sourcepos="2:2-2:199">Implemented several FullTensor functions also for sparse representations (modify(_diag)_elements, fix_slate, resize_dimension,...)</li>
<li data-sourcepos="3:2-3:45">! IndexedTensor*s now considered internal.</li>
<li data-sourcepos="4:2-4:52">! Removed all support for stored IndexedTensor*s.</li>
<li data-sourcepos="5:2-5:50">! Reworked the IndexedTensor* index assignment.</li>
<li data-sourcepos="6:2-6:77">! Changed resize_dimension to allow slate insertion at the first position.</li>
<li data-sourcepos="7:2-7:105">Added TTTensor::random with callback function to manipulate the singular values of all matricisations.</li>
<li data-sourcepos="8:2-8:52">Rewrote the ALS algorithm for better readability.</li>
<li data-sourcepos="9:2-9:88">Added fully functional (multi-site) DMRG and alternating steepest descent algorithms.</li>
<li data-sourcepos="10:2-10:62">Support for low level factorisation calls for Tensor class.</li>
<li data-sourcepos="11:2-11:120">Several bug fixes, including SVD factor handling, SparseTensor summation, Tensor resize_dimension, TN evaluation,....</li>
<li data-sourcepos="12:2-12:28">Added several test cases.</li>
</ul>2016-01-26T15:25:38+01:00Benjamin Huberben.huber@hemio.dehttps://git.hemio.de/xerus/xerus/-/tags/v1.5.0v1.5.0<ul data-sourcepos="1:2-4:112" dir="auto">
<li data-sourcepos="1:2-1:82">Added the ADF algorithm for rank one measurements instead of point evaluations.</li>
<li data-sourcepos="2:2-2:115">Parallelized the ADF algorithm using openmp (note: at this point not compatible with the replacement allocator!)</li>
<li data-sourcepos="3:2-3:106">TT::random no longer uses size_t to measure the size of the tensor and no longer exceeds maximal ranks.</li>
<li data-sourcepos="4:2-4:112">Fixed a bug in the ALS that caused the algorithm to always stop after one sweep if called without operator A.</li>
</ul>2016-01-15T01:38:42+01:00Benjamin Huberben.huber@hemio.dehttps://git.hemio.de/xerus/xerus/-/tags/v1.4.2v1.4.2<ul data-sourcepos="1:2-2:70" dir="auto">
<li data-sourcepos="1:2-1:93">The custom allocator now enforces memory alignment even if the underlying malloc does not.</li>
<li data-sourcepos="2:2-2:70">Cannonicalized TT tensors can no longer exceed the maximal TT rank.</li>
</ul>2015-12-04T01:31:35+01:00Benjamin Huberben.huber@hemio.dehttps://git.hemio.de/xerus/xerus/-/tags/v1.4.1v1.4.1<ul data-sourcepos="1:2-3:93" dir="auto">
<li data-sourcepos="1:2-1:71">Fixed a bug with the recannonicalization of products of TTOperators.</li>
<li data-sourcepos="2:2-2:65">Many improvements for the ADF algorithm for tensor completion.</li>
<li data-sourcepos="3:2-3:93">Added TTTensor::ones and TTOperator::ones to create TTTensors with all entries equal to 1.</li>
</ul>2015-12-04T01:31:26+01:00Benjamin Huberben.huber@hemio.dehttps://git.hemio.de/xerus/xerus/-/tags/v1.4.0v1.4.0<ul data-sourcepos="1:2-17:106" dir="auto">
<li data-sourcepos="1:2-1:48">Added an implementation of the ADF algorithm.</li>
<li data-sourcepos="2:2-2:143">Added a custom allocator to better handle the small allocations of Indices etc. (requires to set the corresponding option in the config.mk).</li>
<li data-sourcepos="3:2-3:75">Reduced overhead (time not spent in blas) of TensorNetwork contractions.</li>
<li data-sourcepos="4:2-4:93">Tensor decomposition now allow arbitrary index orders in the target as well as the source.</li>
<li data-sourcepos="5:2-5:75">Added TN::measure(...) to perform a batch of evaluations of the network.</li>
<li data-sourcepos="6:2-6:70">Added TN::fix_slate(...) to fix a single index to a specific value.</li>
<li data-sourcepos="7:2-7:69">Added TN::reduce_representation() to trace out any full-rank edge.</li>
<li data-sourcepos="8:2-8:110">Added TN::transfer_core(...) and TN::round_edge(...) analogously to TT rounding / core movement operations.</li>
<li data-sourcepos="9:2-9:113">! renamed class::construct_random to class::random and TTOperator::construct_identity to TTOperator::identity.</li>
<li data-sourcepos="10:2-10:36">Added several named constructors:</li>
<li data-sourcepos="11:3-11:78">Tensor::ones(...) to construct tensors filled with all entries equal to 1.</li>
<li data-sourcepos="12:3-12:91">Tensor::identity(...) to construct a (sparse) tensor equal to the TTOperator::identity.</li>
<li data-sourcepos="13:3-13:112">Tensor::dirac(...) to construct a (sparse) tensor with a single 1 at a given position and 0 everywhere else.</li>
<li data-sourcepos="14:3-14:102">Tensor::kronecker(...) to construct a tensor that is 1 when all indices are equal and 0 otherwise.</li>
<li data-sourcepos="15:2-15:117">! misc::approx_equal now uses a relative difference (for consistency with the tensor variant xerus::approx_equal).</li>
<li data-sourcepos="16:2-16:38">Several bugfixes for SparseTensors.</li>
<li data-sourcepos="17:2-17:106">Several minor bugfixes (TN::draw no longer uses temporary files, span-zero indices in assignment, ...).</li>
</ul>2015-12-04T01:31:16+01:00Benjamin Huberben.huber@hemio.dehttps://git.hemio.de/xerus/xerus/-/tags/old-TN-contractionold-TN-contraction<p data-sourcepos="1:1-1:33" dir="auto">meant for use by developers only!</p>2015-12-04T01:27:27+01:00Benjamin Huberben.huber@hemio.dehttps://git.hemio.de/xerus/xerus/-/tags/v1.3.0v1.3.0<ul data-sourcepos="1:2-7:99" dir="auto">
<li data-sourcepos="1:2-1:76">Added several retractions to allow Riemannian algorithms on TT manifolds.</li>
<li data-sourcepos="2:2-2:59">Added the basic (Riemannian) steepest descent algorithm.</li>
<li data-sourcepos="3:2-3:59">Added the (so far untested) CG algorithm for TT tensors.</li>
<li data-sourcepos="4:2-4:85">! New PerformanceData class to log stepcount, time and residual of algorithm runs.</li>
<li data-sourcepos="5:2-5:63">New algorithm to determine the largest entry of a TT tensor.</li>
<li data-sourcepos="6:2-6:110">Fully rewrote the contraction heuristics, now including a local search for more optimal contraction orders.</li>
<li data-sourcepos="7:2-7:99">Several bugfixes for order-zero tensors, TT tensors of more than full rank and other rare cases.</li>
</ul>2015-12-04T01:31:05+01:00Benjamin Huberben.huber@hemio.dehttps://git.hemio.de/xerus/xerus/-/tags/v1.2.1v1.2.1<ul data-sourcepos="1:2-2:48" dir="auto">
<li data-sourcepos="1:2-1:50">entrywise_product for FullTensor and TTNetwork.</li>
<li data-sourcepos="2:2-2:48">TensorNetworks now without extraneous factor.</li>
</ul>2015-12-04T01:30:55+01:00Benjamin Huberben.huber@hemio.dehttps://git.hemio.de/xerus/xerus/-/tags/v1.2.0v1.2.0<ul data-sourcepos="1:2-6:87" dir="auto">
<li data-sourcepos="1:2-1:59">Rank revealing QR decomposition in TTNetwork:move_core.</li>
<li data-sourcepos="2:2-2:59">ALS performance improvement by skipping full-rank nodes.</li>
<li data-sourcepos="3:2-3:89">Added TensorNetwork::draw to get a graphical representation of a given tensor network.</li>
<li data-sourcepos="4:2-4:78">Added algorithm DecompositionALS to approximate a FullTensor by a TTTensor.</li>
<li data-sourcepos="5:2-5:25">New Quick-Start guide.</li>
<li data-sourcepos="6:2-6:87">Several bugfixes, more consistent operator availability and much more documentation.</li>
</ul>2015-12-04T01:30:45+01:00Benjamin Huberben.huber@hemio.dehttps://git.hemio.de/xerus/xerus/-/tags/v1.1.1v1.1.1<ul data-sourcepos="1:2-5:22" dir="auto">
<li data-sourcepos="1:2-1:65">Complete documentation for Tensor, FullTensor and SparseTensor</li>
<li data-sourcepos="2:2-2:45">Added implementation of the projection ALS</li>
<li data-sourcepos="3:2-3:42">Fixed performance bug in logging system</li>
<li data-sourcepos="4:2-4:57">ALS now uses energy functional for convergence testing</li>
<li data-sourcepos="5:2-5:22">Several minor fixes</li>
</ul>2015-12-04T01:30:33+01:00Sebastian Wolfhttps://git.hemio.de/xerus/xerus/-/tags/v1.1.0v1.1.0<ul data-sourcepos="1:2-4:40" dir="auto">
<li data-sourcepos="1:2-1:102">! New representation for the TT tensors: all component tensors are now of degree 3 / 4 (operators).</li>
<li data-sourcepos="2:2-2:77">TT::get_component and ::set_component functions that handle the meta data.</li>
<li data-sourcepos="3:2-3:63">More strict ISO c++ conformity: no names starting with '__'.</li>
<li data-sourcepos="4:2-4:40">More elaborate performance profiling.</li>
</ul>2015-12-04T01:30:18+01:00Benjamin Huberben.huber@hemio.de