chunk_uid
stringlengths
40
40
chunk_type
stringclasses
2 values
chunk_index
int64
0
6.71k
total_chunks
int64
1
6.71k
section_title
stringlengths
1
157
embed_text
stringlengths
1
83.3k
spans
dict
paper_doi
stringlengths
0
63
paper_id_arxiv
stringlengths
9
16
title
stringlengths
7
245
authors
listlengths
1
768
categories
listlengths
1
7
year
int64
2k
2.02k
language
stringclasses
2 values
discipline
stringclasses
8 values
dense_vector
listlengths
1.02k
1.02k
169fb3014f9dd0d20694773163aab461d317bdd4
subsection
76
80
Proof of
\end{align*} \end{}\end{align*} \end{}\begin{} We list some examples of E\left( H^{[n]}_m;t \right): \begin{equation} \begin{array}{|c|c|c| c|} \hline n & m=1 & m=2 & m=3 \\\hline 1 & t^2-1 & 1 & 0 \\ \cline {1-1} 2 & t^4+t^3-t^2-t & t^2+t & 0 \\\cline {1-1} 3 & t^{6}+t^{5}-2 t^3-t^2+t & t^4+2 t^3+t^2-t-1 & 1 \\\cline ...
{ "cite_spans": [] }
1806.03955
Refined Hilbert schemes, E-polynomials, and the number of generators of finite colength ideals in the plane
[ "Yi-Ning Hsiao", "Andras Szenes" ]
[ "math.AG" ]
2,018
en
Mathematics
[ -0.02228672243654728, 0.009305202402174473, -0.02677152492105961, -0.0034398739226162434, -0.02254604734480381, -0.017054453492164612, -0.02414776384830475, 0.047349750995635986, -0.0011145267635583878, 0.004942435305565596, -0.039020832628011703, 0.019434144720435143, 0.018762949854135513, ...
677e59dee14d3b2db70a3215fc4c669c3d72f9c6
subsection
77
80
Proof of
\end{equation} \begin{}[h] \begin{array}{|c|cccc|}\hline n & m=2& m=3&m=4&m=5\\\hline 0 & 0 & 0 & 0 & 0 \\ 1 & 1 & 0 & 0 & 0 \\ 2 & 2 & 0 & 0 & 0 \\ 3 & 2 & 1 & 0 & 0 \\ 4 & 3 & 2 & 0 & 0 \\ 5 & 2 & 5 & 0 & 0 \\ 6 & 4 & 6 & 1 & 0 \\ 7 & 2 & 11 & 2 & 0 \\\hline \end{array} \caption {Examples of \chi \left( B^{[n]}_m\rig...
{ "cite_spans": [] }
1806.03955
Refined Hilbert schemes, E-polynomials, and the number of generators of finite colength ideals in the plane
[ "Yi-Ning Hsiao", "Andras Szenes" ]
[ "math.AG" ]
2,018
en
Mathematics
[ 0.01756073720753193, -0.011557132005691528, -0.03576226532459259, -0.03173443302512169, 0.01299891248345375, -0.022458214312791824, -0.003758928505703807, 0.044672317802906036, 0.006735938601195812, 0.010222149081528187, -0.03118518367409706, 0.019132202491164207, -0.016645321622490883, -0...
76954a7440ee0a9d72f0b4979bc2d24c46a843c7
subsection
78
80
Proof of
Indeed, each strata B^{[1]}_2, B^{[3]}_3 and B^{[6]}_4 contains a single monomial ideal: 1{16pt}\begin{array}{ccc} (:y,~:x) & (:y^2,~:<xy>,~~:<x^2>) & (:<y^3>,~:<y^2x>,~~:<x^2y>,~~~:<x^3>) \\ B^{[1]}_2=\lbrace \mathfrak {m}\rbrace & B^{[3]}_3=\lbrace \langle x^2,xy,y^2\rangle \rbrace & B^{[6]}_4=\lbrace \langle x^3,x^2...
{ "cite_spans": [] }
1806.03955
Refined Hilbert schemes, E-polynomials, and the number of generators of finite colength ideals in the plane
[ "Yi-Ning Hsiao", "Andras Szenes" ]
[ "math.AG" ]
2,018
en
Mathematics
[ -0.019327862188220024, 0.005857852753251791, -0.04140159487724304, -0.028496012091636658, 0.00866474024951458, -0.03221818804740906, 0.029121460393071175, -0.007066797465085983, 0.016658268868923187, 0.0034418697468936443, -0.033682651817798615, -0.009595285169780254, 0.027641741558909416, ...
04056769494789adb9d7f6518d34e3d8ad94a871
subsection
79
80
Proof of
Göttsche, Hilbert schemes of points on surfaces, in Proceedings of the International Congress of Mathematicians, Vol. II (Beijing, 2002), 483–494, Higher Ed. Press, Beijing.L. Göttsche, Hilbert schemes of zero-dimensional subschemes of smooth varieties, Lecture Notes in Mathematics, 1572, Springer-Verlag, Berlin, 1994....
{ "cite_spans": [] }
1806.03955
Refined Hilbert schemes, E-polynomials, and the number of generators of finite colength ideals in the plane
[ "Yi-Ning Hsiao", "Andras Szenes" ]
[ "math.AG" ]
2,018
en
Mathematics
[ -0.04814627394080162, 0.03435532748699188, -0.006189908366650343, -0.006117444951087236, 0.0323110930621624, 0.04985488951206207, 0.019694816321134567, -0.02752087451517582, 0.013508722186088562, -0.0008619342697784305, -0.030007516965270042, 0.02105255424976349, 0.014370656572282314, 0.01...
dfa526767c836147b1e2145bf56444915b80215d
abstract
0
30
Abstract
We suggest an index-free formalism allowing to simplify many computations in Riemann geometry. The main ingredients are forms with values in a Clifford algebra and an action of the group $\mathfrak{sl}_2\times \mathfrak{sl}_2$ on such forms.
{ "cite_spans": [] }
1810.00239
Riemann geometry without indices
[ "Vladimir V. Fock", "Pierre Goussard" ]
[ "math.DG" ]
2,018
en
Mathematics
[ -0.04540373012423515, 0.0428101047873497, -0.027309367433190346, 0.01189254317432642, 0.008536084555089474, -0.007368952501565218, 0.07555083185434341, 0.026958465576171875, 0.031398143619298935, -0.005305493250489235, -0.011015286669135094, -0.009977836161851883, 0.009779499843716621, 0.0...
46231c572ea06cef58bf88ae63a1fdd659984a4c
subsection
1
30
Introduction
Working with Levi-Civita connection, curvature, Weyl and Ricci tensors and in particular deriving Einstein equation out of the Hilbert action is a painful struggle with indices (see for example ). In the present note we suggest a version of Cartan-Palatini formalism allowing to make most of the computations without ind...
{ "cite_spans": [ { "arxiv_id": "", "doi": "", "end": 196, "openalex_id": "", "raw": "B.A.Dubrovin, A.T.Fomenko, and S.P.Novikov, Modern Geometry - Methods and Applications : Part I, 2nd Edition, Springer-Verlag, N.Y., 1992.", "source_ref_id": "40f9ab95e282b33183b3684272d1500d4...
1810.00239
Riemann geometry without indices
[ "Vladimir V. Fock", "Pierre Goussard" ]
[ "math.DG" ]
2,018
en
Mathematics
[ -0.023646119982004166, 0.01185357104986906, -0.045614127069711685, 0.009877976030111313, 0.006422591395676136, 0.021449318155646324, 0.0503738634288311, 0.028329577296972275, 0.044820837676525116, -0.003955004271119833, -0.03194514662027359, -0.00823800265789032, 0.012265470810234547, 0.00...
a77cbf7599094fb2c5d151c649a4c67a012edfea
subsection
2
30
Grassmann and Clifford algebras.
In this section we recall basic notions about Grassmann (exterior) and Clifford algebras and relations between them. Then we define the algebra of Clifford forms and the action of the algebra \mathfrak {sl}_2\times \mathfrak {sl}_2 thereon.A Clifford algebra \mathit {Cl}(\mathrm {V}) of a vector space \mathrm {V} provi...
{ "cite_spans": [] }
1810.00239
Riemann geometry without indices
[ "Vladimir V. Fock", "Pierre Goussard" ]
[ "math.DG" ]
2,018
en
Mathematics
[ -0.012472462840378284, -0.002341924235224724, -0.05443638935685158, 0.012174954637885094, 0.01238855067640543, -0.010199194774031639, 0.012426692992448807, 0.0015390332555398345, 0.03615869954228401, -0.004512208979576826, -0.029949167743325233, -0.014356682077050209, 0.004901258274912834, ...
bc78591d77833e6ab7892bce7a0b6156e347043e
subsection
3
30
Grassmann and Clifford algebras.
The identification of \mathit {Cl}(\mathrm {V}) and \Lambda (\mathrm {V}) allows to consider * also as an automorphism of the Grassmann algebra (as a vector space). The actions v\wedge and v\vdash on the Clifford algebra transferred to the Grassmann algebra are just the usual external and internal multiplication by v.O...
{ "cite_spans": [ { "arxiv_id": "", "doi": "", "end": 1054, "openalex_id": "", "raw": "A.Losev, From Berezin integral to Batalin-Vilkovisky formalism. A mathematical physicist's point of view. in M.Shifman (ed.) Felix Berezin. Life and Death of the Mastermind of Supermathematics. Wor...
1810.00239
Riemann geometry without indices
[ "Vladimir V. Fock", "Pierre Goussard" ]
[ "math.DG" ]
2,018
en
Mathematics
[ -0.023866884410381317, 0.007599557749927044, -0.05383783206343651, 0.030749214813113213, -0.00465434743091464, -0.012276795692741871, 0.01890733279287815, 0.018251147121191025, 0.0405920147895813, 0.03192424774169922, -0.0487104170024395, 0.007710193749517202, -0.005871345289051533, 0.0282...
de58b454406e0948232ba4601a38d54013e10e95
subsection
4
30
Action of the Lie algebra
Consider now the algebra \Omega ^{{\cdot }{\cdot }}=\mathit {Cl}(V)\otimes \Omega (M) of forms on an n-dimensional manifold M with values in the Clifford algebra. Let the vielbein \theta \in \Omega ^{11} defines an isomorphism between the tangent bundle to M and a trivial bundle with fiber \mathrm {V}. One can define t...
{ "cite_spans": [] }
1810.00239
Riemann geometry without indices
[ "Vladimir V. Fock", "Pierre Goussard" ]
[ "math.DG" ]
2,018
en
Mathematics
[ -0.010518213734030724, -0.0058006541803479195, -0.04149317368865013, 0.0006216348265297711, 0.012287775985896587, -0.0459170788526535, 0.01343951653689146, 0.02295853942632675, 0.05680903419852257, -0.004381953738629818, -0.03517766669392586, -0.04082195833325386, 0.03151650354266167, -0.0...
2d34ab8dfcc6d977f1e52730672d6a9e60a0c666
subsection
5
30
Action of the Lie algebra
At one point m\in M a Clifford form takes value in \mathit {Cl}(\mathrm {V})\otimes \Lambda (T_mM) which, using the isomorphism defined by \theta , can be identified with \mathit {Cl}(\mathrm {V})\otimes \Lambda (\mathrm {V}^*).To make explicit computation one can further identify this algebra with the algebra \Lambda ...
{ "cite_spans": [] }
1810.00239
Riemann geometry without indices
[ "Vladimir V. Fock", "Pierre Goussard" ]
[ "math.DG" ]
2,018
en
Mathematics
[ -0.009547112509608269, 0.01956738345324993, -0.050673723220825195, 0.011248955503106117, 0.01303474698215723, -0.01814790815114975, 0.012363167479634285, 0.032541077584028244, 0.06544847041368484, 0.02260475419461727, -0.028084231540560722, -0.016728434711694717, 0.016468960791826248, 0.01...
6b25542a804a1ad6aeccb8e64d7cba09415d136c
subsection
6
30
Action of the Lie algebra
The property 3 follows from the computation similar to (REF ).2E^{\prime }x&=\theta x-(-1)^{p+q}x\theta =\xi _ia\otimes \xi ^i\alpha -(-1)^{p+q} a\xi _i\otimes \alpha \xi ^i=\\&=\xi _ia\otimes \xi ^i\alpha -(-1)^p a\xi _i\otimes \xi ^i\alpha =\xi _i\vdash a\otimes \xi ^i\alpha =\eta _{ij}\frac{\partial }{\partial \xi _...
{ "cite_spans": [] }
1810.00239
Riemann geometry without indices
[ "Vladimir V. Fock", "Pierre Goussard" ]
[ "math.DG" ]
2,018
en
Mathematics
[ -0.006280547473579645, 0.06138605624437332, -0.017826681956648827, 0.014705485664308071, -0.02649582177400589, -0.0059905587695539, 0.06263758987188339, 0.04606243595480919, 0.033882904797792435, 0.017918257042765617, -0.0190629493445158, 0.01985660381615162, -0.006944469176232815, 0.03904...
0e16b28825dbccb5d81130a843c6e695d772a68d
subsection
7
30
Supertrace.
A supertrace of a Clifford algebra is a linear function on \mathit {Cl}(\mathrm {V}) satisfying the identity \operatorname{str}(ab)=(-1)^{\deg a\deg b}\operatorname{str}(ba) and normalized by the condition that \operatorname{str}\operatorname{{v}}=1. If n is even then the Clifford algebra has a unique representation an...
{ "cite_spans": [] }
1810.00239
Riemann geometry without indices
[ "Vladimir V. Fock", "Pierre Goussard" ]
[ "math.DG" ]
2,018
en
Mathematics
[ -0.0332338772714138, -0.022277988493442535, -0.05883830413222313, 0.019973894581198692, -0.00003266357452957891, -0.019027844071388245, 0.018722666427493095, -0.0036659499164670706, 0.028763020411133766, 0.010803299024701118, -0.015136824920773506, -0.015014753676950932, -0.00787359010428190...
836afab0abbb36a6bc4e2a8465374daf570bb59c
subsection
8
30
The action of the group
The invertible elements of g\in \mathit {Cl}(\mathrm {V}) such that g^{-1}\mathit {Cl}^1(\mathrm {V})g\subset \mathit {Cl}^1(\mathrm {V}) form a Lie group denoted by Pin(\mathrm {V}). The Lie algebra of this group is \mathit {Cl}^2(\mathrm {V}) with respect to the commutator. It is isomorphic to the Lie algebra \mathfr...
{ "cite_spans": [] }
1810.00239
Riemann geometry without indices
[ "Vladimir V. Fock", "Pierre Goussard" ]
[ "math.DG" ]
2,018
en
Mathematics
[ -0.06223204731941223, -0.004562869668006897, -0.031314410269260406, 0.02450825646519661, 0.015290954150259495, -0.028674354776740074, 0.018938198685646057, 0.008156701922416687, -0.007508133538067341, 0.015886111184954643, -0.054052457213401794, -0.014917073771357536, -0.05866110697388649, ...
912a87826abffd25dca46b54008b50a70725f3b3
subsection
9
30
Use of the formalism
Let M be an n-dimensional manifold and \mathrm {V} an n-dimensional vector space with a nondegenerate quadratic form \eta . Let we are given two Clifford algebra valued 1-forms \theta \in \Omega ^{11} and \omega \in \Omega ^{21} such that the form \theta is nondegenerate in the sense that it induces an isomorphism at e...
{ "cite_spans": [] }
1810.00239
Riemann geometry without indices
[ "Vladimir V. Fock", "Pierre Goussard" ]
[ "math.DG" ]
2,018
en
Mathematics
[ -0.025593668222427368, 0.027364011853933334, -0.03284291923046112, -0.011430922895669937, -0.01762712374329567, -0.06556374579668045, 0.029058046638965607, 0.013208896853029728, 0.05646784231066704, 0.013094434514641762, -0.029180139303207397, 0.005669676698744297, 0.0007740482105873525, -...
da77f5727bef3359a0c009d2d28cf3df62358a52
subsection
10
30
Gauge group action.
A pair \theta ,\omega defines the same metric and connection as a pair \theta ^{\prime },\omega ^{\prime } if and only if there exists a gauge transformation relating them, namely if there exists a function g on M with values in Pin(\mathrm {V})\subset \mathit {Cl}(\mathrm {V}) such that \theta ^{\prime }=g^{-1}\theta ...
{ "cite_spans": [] }
1810.00239
Riemann geometry without indices
[ "Vladimir V. Fock", "Pierre Goussard" ]
[ "math.DG" ]
2,018
en
Mathematics
[ -0.03245070204138756, 0.037854064255952835, -0.02729155868291855, 0.028070010244846344, 0.013088652864098549, -0.038678307086229324, 0.025749921798706055, 0.038311976939439774, 0.009921427816152573, -0.0001485829270677641, -0.02562781237065792, -0.02025497704744339, 0.03669402003288269, -0...
b7d1aff4a7cc8b8b98004b8e347310734518c367
subsection
11
30
Curvature.
The connection \nabla can be extended in a standard way to \Omega ^{pq} by the formula \nabla e \alpha = (\nabla e) \alpha + ed\alpha for e\in \Omega ^{p0} and \alpha \in \Omega ^{0q}. It can be therefore written as \nabla x = dx + \omega x -(-1)^q x\omega for any x\in \Omega ^{pq}. This allows to compute \nabla ^2 x =...
{ "cite_spans": [] }
1810.00239
Riemann geometry without indices
[ "Vladimir V. Fock", "Pierre Goussard" ]
[ "math.DG" ]
2,018
en
Mathematics
[ -0.034795764833688736, 0.03403269872069359, -0.015978582203388214, -0.005139242857694626, -0.03842795640230179, 0.006188457366079092, 0.052346259355545044, 0.024250207468867302, -0.012865276075899601, -0.005719172302633524, -0.03507046774029732, 0.026005256921052933, -0.028752289712429047, ...
a8e31c5af20c88044da3007487d6b4dd6df44181
subsection
12
30
Torsion.
The torsion can be defined as an element t\in \Omega ^{12} such that for any two vector fields X and Y on M we have i_Yi_Xt = i_X\nabla i_Y\theta -i_Y\nabla i_X\theta -i_{[X,Y]}\theta . Taking into account the identity i_Xi_Yd\alpha =i_Ydi_X\alpha -i_Xdi_Y\alpha +i_{[X,Y]}\alpha valid for any 1-form \alpha one can easi...
{ "cite_spans": [] }
1810.00239
Riemann geometry without indices
[ "Vladimir V. Fock", "Pierre Goussard" ]
[ "math.DG" ]
2,018
en
Mathematics
[ 0.00031209737062454224, -0.009376750327646732, -0.03134230151772499, 0.020966077223420143, 0.017914246767759323, -0.03637782111763954, 0.008926604874432087, 0.06475985050201416, 0.04473983868956566, -0.014389381743967533, -0.03265459090471268, -0.0406503863632679, 0.010238892398774624, -0....
ae36372a963a824b16fa18130fa960765d6a9694
subsection
13
30
Bianchi identities.
Computing the covariant derivative of the torsion one gets\nabla t= dt+\omega t-t\omega =R\theta -\theta R=-2E^{\prime }R.In particular this identity implies that if the torsion vanishes we have E^{\prime }R=0. Since H^{\prime }R=0 we have F^{\prime }R=0 and therefore the curvature form is invariant with respect to the...
{ "cite_spans": [ { "arxiv_id": "", "doi": "10.48550/arxiv.math/0204191", "end": 619, "openalex_id": "https://openalex.org/W1552534450", "raw": "P.Ševera, A remark on the symmetries of the Riemann curvature tensor., arxiv:math/0204191.", "source_ref_id": "192d05d58f4444b4de8b44...
1810.00239
Riemann geometry without indices
[ "Vladimir V. Fock", "Pierre Goussard" ]
[ "math.DG" ]
2,018
en
Mathematics
[ -0.0499664805829525, 0.020847346633672714, 0.011232538148760796, 0.01947380229830742, -0.0008322341600432992, 0.028417101129889488, 0.05176734924316406, 0.06892138719558716, 0.028081346303224564, 0.006119901780039072, -0.012728175148367882, -0.009683486074209213, 0.006688396446406841, -0.0...
4994e3f1be32652719b42137c626125bcf5cf21b
subsection
14
30
Conformal transformations and the Weyl tensor.
Let (\theta ,\omega ) be a pair of Clifford forms with vanishing torsion. Let \tilde{\theta }=e^\phi \theta be a conformal transformation of the form \theta . This transformation corresponds to a conformal transformation of the metric. Compute now the induced transformation of the form \omega and of the curvature R.Pro...
{ "cite_spans": [] }
1810.00239
Riemann geometry without indices
[ "Vladimir V. Fock", "Pierre Goussard" ]
[ "math.DG" ]
2,018
en
Mathematics
[ -0.05194837227463722, 0.020480234175920486, 0.005242146551609039, -0.005520659498870373, 0.011689910665154457, -0.037725139409303665, 0.0011350353015586734, 0.033513110131025314, 0.06089130789041519, 0.007767838425934315, -0.03919019550085068, -0.014581865631043911, 0.014917607419192791, -...
fee1407b5867520cdd3c0935bb84ce1b56818b23
subsection
15
30
Conformal transformations and the Weyl tensor.
The torsion and the curvature can be computed directly:&e^{-\phi }(d\tilde{\theta }+\tilde{\theta }\tilde{\omega }+\tilde{\omega }\tilde{\theta })=e^{-\phi }d(e^{\phi }\theta )+\theta \omega +\omega \theta +\\&+\theta (\theta \varepsilon -\varepsilon \theta )+(\theta \varepsilon -\varepsilon \theta )\theta =d\theta +\t...
{ "cite_spans": [] }
1810.00239
Riemann geometry without indices
[ "Vladimir V. Fock", "Pierre Goussard" ]
[ "math.DG" ]
2,018
en
Mathematics
[ -0.010358802042901516, 0.03756019473075867, -0.023539958521723747, 0.004237345419824123, -0.008482318371534348, -0.003863574005663395, 0.004759862087666988, 0.04293029382824898, 0.03298340365290642, 0.0340513214468956, -0.04042831435799599, 0.0006984372739680111, -0.0046683261170983315, -0...
abbffc0c0511ada8b9f5090abf6edd444f6fee83
subsection
16
30
Hilbert action.
On the space of pairs \theta ,\omega define the Hilbert functional:S(\theta , \omega )=\operatorname{str}\int _M \theta ^{n-2}(d\omega +\omega ^2)=\operatorname{str}\int _M E^{n-2}R.Proposition 3 The Hilbert functional is gauge invariant. The variation of this functional is\frac{\delta S}{\delta \theta } =&(-1)^{(n^2+n...
{ "cite_spans": [] }
1810.00239
Riemann geometry without indices
[ "Vladimir V. Fock", "Pierre Goussard" ]
[ "math.DG" ]
2,018
en
Mathematics
[ -0.08210186660289764, 0.029559724032878876, -0.022371996194124222, 0.032505013048648834, 0.004520943854004145, -0.0026133726350963116, -0.0036835200153291225, 0.04691099375486374, 0.03793777897953987, 0.0016614946071058512, -0.03277970477938652, 0.0011474040802568197, 0.023577580228447914, ...
62b8bf50c5dcca020700997e5753e49f4bd10bf9
subsection
17
30
Hilbert action.
Computing the variation of the Hilbert functional\delta S=\operatorname{str}\int _{\partial M} \theta ^{n-2}\delta \omega + \operatorname{str}\int _M \frac{\delta S}{\delta \theta }\delta \theta + \operatorname{str}\int _M \frac{\delta S}{\delta \omega }\delta \omegaone gets\frac{\delta S}{\delta \theta } =\Pi ^{n-1}\l...
{ "cite_spans": [] }
1810.00239
Riemann geometry without indices
[ "Vladimir V. Fock", "Pierre Goussard" ]
[ "math.DG" ]
2,018
en
Mathematics
[ -0.07274455577135086, 0.026440083980560303, -0.030025437474250793, 0.017850488424301147, -0.018201395869255066, -0.0006207622354850173, -0.017972543835639954, 0.030650967732071877, 0.03246653079986572, 0.011671474203467369, -0.03225293383002281, 0.026501111686229706, -0.004710545763373375, ...
c264bb50e21e8018f22deab6872b1852908b2600
subsection
18
30
Special case
The dimension 4 of the manifold M is special in particular since in this case HR=0 and the equation of motion (Einstein equation) amounts to ER=0. Together with the Bianchi identity E^{\prime }R=0 it implies that the curvature R is invariant under the action of both \mathfrak {sl}_2 algebras. In this dimension the acti...
{ "cite_spans": [ { "arxiv_id": "", "doi": "10.1098/rspa.1978.0143", "end": 674, "openalex_id": "https://openalex.org/W2032317001", "raw": "M.F.Atiyah, N.J.Hitchin, I.M.Singer: Self-duality in four-dimensional Riemannian geometry, Proc.Roy.Soc.Lond.A 362 (1978), 425–461.", "sou...
1810.00239
Riemann geometry without indices
[ "Vladimir V. Fock", "Pierre Goussard" ]
[ "math.DG" ]
2,018
en
Mathematics
[ -0.030204789713025093, 0.050036218017339706, 0.01835169829428196, 0.009008044376969337, -0.01577361300587654, 0.022226454690098763, 0.045215655118227005, 0.0480225645005703, 0.01417184341698885, 0.017360126599669456, -0.026009680703282356, -0.013042978011071682, 0.016292281448841095, 0.003...
3d0f0d81550b18d6548f2f3bb13488a572c4f702
subsection
19
30
Complex structure
In this section we generalize the construction for the case where the space \mathrm {V} is provided with a complex structure J compatible with the metric. In order to simplify notation we rename this space, vielbein, connection, torsion, curvature etc. into \mathrm {V}^{\mathbb {R}}, \theta ^{\mathbb {R}} t^{\mathbb {R...
{ "cite_spans": [] }
1810.00239
Riemann geometry without indices
[ "Vladimir V. Fock", "Pierre Goussard" ]
[ "math.DG" ]
2,018
en
Mathematics
[ -0.025342250242829323, 0.04369669035077095, -0.019269874319434166, 0.011374261230230331, 0.059442147612571716, -0.01939193345606327, 0.010825000703334808, 0.017118606716394424, 0.03457287326455116, 0.010603771544992924, -0.00033875808003358543, -0.012007435783743858, 0.021192286163568497, ...
18b9e2201ae303bcdf72a15b2460ec5d18a02a8e
subsection
20
30
Complex structure
In this case one can define a decomposition of the differential d=\partial +\bar{\partial } with the property \partial :\Omega ^{p\bar{p}q\bar{q}}\rightarrow \Omega ^{p\bar{p}(q+1)\bar{q}} and \bar{\partial }:\Omega ^{p\bar{p}q\bar{q}}\rightarrow \Omega ^{p\bar{p}q(\bar{q}+1)}.
{ "cite_spans": [] }
1810.00239
Riemann geometry without indices
[ "Vladimir V. Fock", "Pierre Goussard" ]
[ "math.DG" ]
2,018
en
Mathematics
[ 0.014487840235233307, 0.025466250255703926, -0.00978062767535448, -0.01061983872205019, -0.03979387879371643, -0.02139226160943508, 0.050444234162569046, 0.05285505950450897, 0.01468619890511036, -0.00276939757168293, -0.011169140227138996, -0.009864548221230507, -0.026198653504252434, 0.0...
09a5abb8ea761b380cf2a827c4b058b846c29764
subsection
21
30
Action of the algebra
Define the operators2E_0x&=\theta x+(-1)^{(p+\bar{p}+q+\bar{q})}x\theta ,& 2F_1x&=\theta x-(-1)^{(p+\bar{p}+q+\bar{q})}x\theta \\ 2F_2x&=\bar{\theta } x+(-1)^{(p+\bar{p}+q+\bar{q})}x\bar{\theta }, & 2E_3&=\bar{\theta } x-(-1)^{(p+\bar{p}+q+\bar{q})}x\bar{\theta }Observe that E_3=*_1^{-1}F_2*_1 and F_1=*_1^{-1}E_0*_1. D...
{ "cite_spans": [] }
1810.00239
Riemann geometry without indices
[ "Vladimir V. Fock", "Pierre Goussard" ]
[ "math.DG" ]
2,018
en
Mathematics
[ -0.0011441675014793873, 0.012364637106657028, -0.05754399672150612, 0.028634699061512947, 0.014317349530756474, 0.0017725061625242233, 0.02948901057243347, 0.05928312987089157, 0.01261635310947895, 0.010541596449911594, -0.03874913975596428, -0.008222750388085842, -0.01192985288798809, -0....
f2fd93ca93d27c2ad82e15d5f00f40492c176cc3
subsection
22
30
Action of the algebra
These operators have degrees\deg {E_0}&=(1,0,1,0),& \deg {E_1}&=(0,1,-1,0),& \deg {E_2}&=(0,-1,0,-1),\\ \deg {E_3}&=(-1,0,0,1), & \deg {F_0}&=(-1,0,-1,0),& \deg {F_1}&=(0,-1,1,0), \\ \deg {F_2}&=(0,1,0,1),&\deg {F_3}&=(1,0,0,-1), &\deg H_i&=(0,0,0,0).Proposition 4 The operators \lbrace E_i,F_i,H_i\rbrace generate the a...
{ "cite_spans": [] }
1810.00239
Riemann geometry without indices
[ "Vladimir V. Fock", "Pierre Goussard" ]
[ "math.DG" ]
2,018
en
Mathematics
[ -0.02053518407046795, -0.0229151900857687, -0.02990264631807804, 0.003022684482857585, 0.03307598829269409, 0.004592191893607378, -0.0033907464239746332, 0.022991472855210304, 0.040216006338596344, -0.010473554022610188, -0.05104808881878853, -0.02482224814593792, -0.00049392762593925, 0.0...
131e74d978e2b6df4faefa263dacf53e46a7b3ac
subsection
23
30
Action of the algebra
Taking into account that*^{-1}\frac{\partial }{\partial \xi _i}*=\eta ^{i\bar{j}}\bar{\xi }_j, && *^{-1}\frac{\partial }{\partial \bar{\xi }_i}*=\eta ^{\bar{i}j}\xi _j, && *^{-1}\xi _i* =\eta _{\bar{j}i} \frac{\partial }{\partial \bar{\xi }_j}, && *^{-1}\bar{\xi }_i* =\eta _{j\bar{i}} \frac{\partial }{\partial \xi _j},...
{ "cite_spans": [] }
1810.00239
Riemann geometry without indices
[ "Vladimir V. Fock", "Pierre Goussard" ]
[ "math.DG" ]
2,018
en
Mathematics
[ -0.010361108928918839, 0.04550342634320259, -0.029664205387234688, 0.03482187166810036, -0.029664205387234688, -0.022187115624547005, 0.006111375521868467, 0.059358928352594376, 0.013000979088246822, 0.02128681354224682, -0.03582898899912834, 0.02945057488977909, -0.018845316022634506, 0.0...
b4009e5f53d8c2d05f316ca9a7fd65bd748df8bf
subsection
24
30
Action of the algebra
To prove () compute the commutators:[E_0,E_1]&=\eta ^{i\bar{j}}\,\xi _i\bar{\xi }_j\otimes 1,& [E_1,E_2]&=1\otimes \eta ^{i\bar{j}}\,\frac{\partial ^2}{\partial \xi ^i\partial \bar{\xi }^j},\\ [E_2,E_3]&=-\eta _{i\bar{j}}\,\frac{\partial ^2}{\partial \xi _i\partial \bar{\xi }_j}\otimes 1,& [E_3,E_0]&=-1\otimes \eta _{i...
{ "cite_spans": [] }
1810.00239
Riemann geometry without indices
[ "Vladimir V. Fock", "Pierre Goussard" ]
[ "math.DG" ]
2,018
en
Mathematics
[ -0.029549507424235344, 0.02941220998764038, -0.03182254731655121, 0.0005067618330940604, -0.01873350329697132, 0.002171972766518593, 0.00931335799396038, 0.021235371008515358, 0.004839742556214333, 0.016124848276376724, -0.012524597346782684, -0.010137143544852734, -0.012532224878668785, 0...
fa0571e28a79afbb9a5e5dcaed1f70deda556520
subsection
25
30
Kähler condition.
The Lie algebra of the unitary group (the group preserving both the complex structure J and the scalar product \eta ) is embedded into the Clifford algebra as \mathit {Cl}^{11}(\mathrm {V}^\mathbb {C}). The connection preserving unitary structure is therefore a form \omega ^{\mathbb {R}}=\omega +\bar{\omega } with \ome...
{ "cite_spans": [] }
1810.00239
Riemann geometry without indices
[ "Vladimir V. Fock", "Pierre Goussard" ]
[ "math.DG" ]
2,018
en
Mathematics
[ -0.025439292192459106, 0.05130587890744209, -0.04535427689552307, 0.00364917260594666, 0.017076525837183, -0.05133640021085739, 0.024340534582734108, 0.018907789140939713, 0.007889690808951855, 0.009171574376523495, -0.03369523584842682, -0.011872687377035618, -0.02066274918615818, -0.0028...
445c4365ad665780fe15e1e44e4e232def5d5ded
subsection
26
30
Kähler condition.
The proof that \bar{\partial }w=0 is analogous and thus the form w is necessarily closed.Conversely, the condition dw=0, the relation t^{0111}=0 and the invertibility of E_3:\Omega ^{1110}\rightarrow \Omega ^{0111} imply that the condition t^{1020}=0 is also satisfied. Indeed the condition t^{0111}=0 implies \omega = -...
{ "cite_spans": [] }
1810.00239
Riemann geometry without indices
[ "Vladimir V. Fock", "Pierre Goussard" ]
[ "math.DG" ]
2,018
en
Mathematics
[ -0.010001526214182377, 0.05126640200614929, -0.027921879664063454, -0.0119011290371418, -0.011725664138793945, -0.04162343591451645, 0.047879159450531006, 0.032346658408641815, 0.009787916205823421, 0.0062099481001496315, -0.01309124194085598, 0.024382056668400764, 0.004642203450202942, 0....
7b006fc73591300686162c7c58affc5ca5c9aebd
subsection
27
30
Bianchi identity in the Kähler case.
In the Kahler case the first Bianchi identity implies that only the component R\in \Omega ^{1111} of the curvature R^\mathbb {R} is nonzero and that it is invariant under a subalgebra \widehat{\mathfrak {sl}}_2\times \widehat{\mathfrak {sl}}_2 of \widehat{\mathfrak {sl}}_4.Proposition 6 The curvature is given by R^{\ma...
{ "cite_spans": [] }
1810.00239
Riemann geometry without indices
[ "Vladimir V. Fock", "Pierre Goussard" ]
[ "math.DG" ]
2,018
en
Mathematics
[ -0.042675863951444626, 0.03357900679111481, -0.017033712938427925, 0.017461081966757774, -0.022665828466415405, 0.016163712367415428, 0.05165061354637146, 0.054459039121866226, -0.005128429736942053, -0.006906590890139341, -0.03354847803711891, -0.034189533442258835, 0.011714493855834007, ...
3b6448cbebdef77c274e697aa49550efd31bc077
subsection
28
30
Einstein equation in the Kähler case.
In the Kähler case the Einstein equation is equivalent to the equations F_0R=0 and E_2R=0. In four real dimension it is equivalent to the full \widehat{\mathfrak {sl}}_4 invariance of the curvature. Indeed, the Einstein equation reads as FR^\mathbb {R}=0. In the complex case it amounts to (F_0+E_2)R=0. Since the two te...
{ "cite_spans": [] }
1810.00239
Riemann geometry without indices
[ "Vladimir V. Fock", "Pierre Goussard" ]
[ "math.DG" ]
2,018
en
Mathematics
[ -0.06505528837442398, 0.026897858828306198, 0.014768648892641068, 0.001094680279493332, -0.013426044024527073, -0.015150070190429688, 0.05153769999742508, 0.026501180604100227, 0.025906162336468697, 0.016675757244229317, -0.035853639245033264, 0.007632248569279909, 0.008025112561881542, 0....
12521fbb0629b615b89b52d07a05a45438deee4b
subsection
29
30
Conclusion.
First of all we hope that the developed formalism allows to simplify learning Riemann differential geometry for students as well as to better understand the logic of the theory. Emergence of the affine group \widehat{\mathfrak {sl}}_4 seems mysterious for us and requires better understanding of its consequences. In the...
{ "cite_spans": [] }
1810.00239
Riemann geometry without indices
[ "Vladimir V. Fock", "Pierre Goussard" ]
[ "math.DG" ]
2,018
en
Mathematics
[ -0.04081081226468086, 0.05790434777736664, -0.005341729614883661, 0.002415224676951766, -0.006822151597589254, -0.025396108627319336, 0.04670197516679764, 0.026891792193055153, 0.019596515223383904, 0.013415371999144554, -0.01704774796962738, 0.009897461161017418, -0.037208959460258484, 0....
623027b1c982101eb2e9716e7a277a4e0c039338
abstract
0
29
Abstract
The way developers edit day-to-day code tends to be repetitive, often using existing code elements. Many researchers have tried to automate repetitive code changes by learning from specific change templates which are applied to limited scope. The advancement of deep neural networks and the availability of vast open-sou...
{ "cite_spans": [] }
10.1109/TSE.2020.3020502
1810.00314
CODIT: Code Editing with Tree-Based Neural Models
[ "Saikat Chakraborty", "Yangruibo Ding", "Miltiadis Allamanis", "Baishakhi Ray" ]
[ "cs.SE" ]
2,018
en
Computer Science
[ -0.06432358175516129, -0.00948223564773798, -0.017545567825436592, 0.016935287043452263, 0.013281232677400112, -0.0625537633895874, 0.002977025229483843, 0.005339955445379019, -0.011358847841620445, 0.01794224977493286, -0.03402314335107803, -0.035304732620716095, -0.03829510882496834, 0.0...
772c39b029699751d71596184ac5ca8d36126fa5
subsection
1
29
0pt
8pt plus 4pt minus 2pt4pt plus 2pt minus 2ptplaintop tablecustomjava belowcaptionskip=breaklines=true, language=java, showstringspaces=false, basicstyle=, keywordstyle=, commentstyle=, identifierstyle=, belowskip=-2pt, aboveskip=-2ptcodit belowcaptionskip=breaklines=true, language=java, showstringspaces=false, basicsty...
{ "cite_spans": [] }
10.1109/TSE.2020.3020502
1810.00314
CODIT: Code Editing with Tree-Based Neural Models
[ "Saikat Chakraborty", "Yangruibo Ding", "Miltiadis Allamanis", "Baishakhi Ray" ]
[ "cs.SE" ]
2,018
en
Computer Science
[ -0.08143474906682968, -0.018054218962788582, -0.030782213434576988, 0.022831032052636147, 0.005879448726773262, -0.028996631503105164, 0.017001183703541756, -0.021320154890418053, -0.01767268404364586, 0.017840558663010597, -0.027363663539290428, -0.013384235091507435, -0.024265602231025696,...
0447d2363a0ebcc6e04ab32c8e47eacb839fa204
subsection
2
29
Introduction
Developers edit source code to add new features, fix bugs, or maintain existing functionality (e.g., API updates, refactoring, etc.) all the time. Recent research has shown that these edits are often repetitive , , . Moreover, the code components (e.g., token, sub-trees, etc.) used to build the edits are often taken fr...
{ "cite_spans": [ { "arxiv_id": "", "doi": "10.1145/1806799.1806847", "end": 216, "openalex_id": "https://openalex.org/W2143051121", "raw": "T. T. Nguyen, H. A. Nguyen, N. H. Pham, J. Al-Kofahi, and T. N. Nguyen, “Recurring bug fixes in object-oriented programs,” in Proceedings of th...
10.1109/TSE.2020.3020502
1810.00314
CODIT: Code Editing with Tree-Based Neural Models
[ "Saikat Chakraborty", "Yangruibo Ding", "Miltiadis Allamanis", "Baishakhi Ray" ]
[ "cs.SE" ]
2,018
en
Computer Science
[ -0.04159925878047943, 0.013055086135864258, -0.021104833111166954, 0.003000533441081643, 0.01113230362534523, -0.08167249709367752, -0.0075995707884430885, 0.0037406522314995527, 0.006077367812395096, 0.005684418138116598, -0.037295885384082794, -0.01643521711230278, -0.017366087064146996, ...
8e8ead4a9d56c2eac65bee6cb93bb4785e17d244
subsection
3
29
Introduction
In the second step, the model concretizes the previously generated code fragment by predicting the tokens conditioned on the AST that was generated in the first step: given the type of each leaf node in the syntax tree, our model suggests concrete tokens of the correct type while respecting scope information. We combin...
{ "cite_spans": [ { "arxiv_id": "", "doi": "10.5555/2820518.2820526", "end": 605, "openalex_id": "https://openalex.org/W1988346254", "raw": "B. Ray, M. Nagappan, C. Bird, N. Nagappan, and T. Zimmermann, “The uniqueness of changes: Characteristics and applications,” ser. MSR '15. ACM,...
10.1109/TSE.2020.3020502
1810.00314
CODIT: Code Editing with Tree-Based Neural Models
[ "Saikat Chakraborty", "Yangruibo Ding", "Miltiadis Allamanis", "Baishakhi Ray" ]
[ "cs.SE" ]
2,018
en
Computer Science
[ -0.0498354434967041, 0.01571662724018097, -0.02594006434082985, 0.013717716559767723, 0.007724798750132322, -0.043121542781591415, 0.012458860874176025, -0.007232700474560261, -0.009246869944036007, 0.03466813266277313, -0.004726432263851166, -0.03475968539714813, -0.02050791122019291, 0.0...
f41a6938f7934ed514660e8fb3da56558275434f
subsection
4
29
Background
Modeling Code Changes. Generating source code using machine learning models has been explored in the past , , , . These methods model a probability distribution p(c|\kappa ) where c is the generated code and \kappa is any contextual information upon which the generated code is conditioned. In this work, we generate cod...
{ "cite_spans": [ { "arxiv_id": "", "doi": "10.5555/2337223.2337322", "end": 113, "openalex_id": "https://openalex.org/W2142403498", "raw": "A. Hindle, E. T. Barr, Z. Su, M. Gabel, and P. Devanbu, “On the naturalness of software,” in 2012 34th International Conference on Software Eng...
10.1109/TSE.2020.3020502
1810.00314
CODIT: Code Editing with Tree-Based Neural Models
[ "Saikat Chakraborty", "Yangruibo Ding", "Miltiadis Allamanis", "Baishakhi Ray" ]
[ "cs.SE" ]
2,018
en
Computer Science
[ -0.02309531718492508, -0.0015302173560485244, -0.02060883305966854, 0.022286828607320786, 0.006197141949087381, -0.07419790327548981, 0.001881070900708437, 0.02561230957508087, 0.01094510592520237, -0.0070475805550813675, -0.03798371180891991, -0.013759561814367771, -0.00730309309437871, 0...
92fa67a54e898e9002d5a4a2e9b70822d853f6b2
subsection
5
29
Motivating Example
fig:motiv illustrates an example of our approach. Here, the original code fragment [rgb]1.00,0.00,0.00return super.equals(object) is edited to [rgb]0.0,0.6,0return object == this. Codit takes these two code fragments along with their context, for training. While suggesting changes, i.e., during test time, Codit takes a...
{ "cite_spans": [ { "arxiv_id": "", "doi": "", "end": 2767, "openalex_id": "https://openalex.org/W2964308564", "raw": "D. Bahdanau, K. Cho, and Y. Bengio, “Neural machine translation by jointly learning to align and translate,” arXiv preprint arXiv:1409.0473, 2014.", "source_re...
10.1109/TSE.2020.3020502
1810.00314
CODIT: Code Editing with Tree-Based Neural Models
[ "Saikat Chakraborty", "Yangruibo Ding", "Miltiadis Allamanis", "Baishakhi Ray" ]
[ "cs.SE" ]
2,018
en
Computer Science
[ -0.04984438419342041, -0.026127859950065613, -0.010889151133596897, 0.022022489458322525, 0.004376263823360205, -0.05243885889649391, 0.0008818343048915267, 0.006936396937817335, -0.004361002240329981, 0.026387305930256844, -0.04718886688351631, -0.03186621889472008, -0.03250720724463463, ...
14ea526a2b32edcfbfba607b55d87ae3fb46e1ba
subsection
6
29
Motivating Example
However, in contrast to traditional seq2seq where the generation of each token is only conditioned on the previously generated and source tokens, we additionally condition on the token type that has been predicted by the tree model and generate only tokens that are valid for that toke type. fig:motivtoken shows this st...
{ "cite_spans": [] }
10.1109/TSE.2020.3020502
1810.00314
CODIT: Code Editing with Tree-Based Neural Models
[ "Saikat Chakraborty", "Yangruibo Ding", "Miltiadis Allamanis", "Baishakhi Ray" ]
[ "cs.SE" ]
2,018
en
Computer Science
[ -0.03860551863908768, -0.014305404387414455, 0.004451842047274113, 0.011657951399683952, 0.004219140857458115, -0.03906329348683357, 0.02739008143544197, 0.04449553042650223, -0.019455350935459137, -0.0006113176350481808, -0.02975524216890335, -0.02729852683842182, -0.002901136176660657, 0...
05ae9f53620e78e1b008361d04e3f2c5cf482bd7
subsection
7
29
Tree-based Neural Translation Model
We decompose the task of predicting code changes in two stages: First, we learn and predict the structure (syntax tree) of the edited code. Then, given the predicted tree structure, we concretize the code. We factor the generation process as\vspace{-14.22636pt} P(c_{n}|c_{p}) = P(c_n|t_n, c_p) P(t_n|t_p) P(t_p|c_p)and ...
{ "cite_spans": [] }
10.1109/TSE.2020.3020502
1810.00314
CODIT: Code Editing with Tree-Based Neural Models
[ "Saikat Chakraborty", "Yangruibo Ding", "Miltiadis Allamanis", "Baishakhi Ray" ]
[ "cs.SE" ]
2,018
en
Computer Science
[ -0.0033725022803992033, 0.0016290254425257444, -0.020738599821925163, 0.03345033898949623, -0.0037807123735547066, -0.015176260843873024, -0.0002494087675586343, 0.016770951449871063, 0.002527469303458929, 0.03924921154975891, -0.02269190549850464, -0.02028079517185688, -0.045780573040246964...
fba533e581ad4d8f3f724896940093f73cd2a377
subsection
8
29
Tree Translation Model (
The goal of {M}_{tree} is to model the probability distribution of a new tree (t_n) given a previous version of the tree (t_p). For any meaningful code the generated tree is syntactically correct. We represent the tree as a sequence of grammar rule generations following the CFG of the underlying programming language. T...
{ "cite_spans": [ { "arxiv_id": "", "doi": "", "end": 2607, "openalex_id": "https://openalex.org/W2964308564", "raw": "D. Bahdanau, K. Cho, and Y. Bengio, “Neural machine translation by jointly learning to align and translate,” arXiv preprint arXiv:1409.0473, 2014.", "source_re...
10.1109/TSE.2020.3020502
1810.00314
CODIT: Code Editing with Tree-Based Neural Models
[ "Saikat Chakraborty", "Yangruibo Ding", "Miltiadis Allamanis", "Baishakhi Ray" ]
[ "cs.SE" ]
2,018
en
Computer Science
[ -0.03664994239807129, -0.020033877342939377, -0.041746146976947784, 0.00045368989231064916, 0.0050008404068648815, -0.02976854145526886, 0.0016850657993927598, 0.021956397220492363, 0.02555730752646923, 0.008941243402659893, -0.037504397332668304, 0.0016364306211471558, -0.00904042087495327,...
e8a5d2a4f49c64366b6288e1affbb28439150dcc
subsection
9
29
Tree Translation Model (
At a given decoding step k the decoder LSTM changes its internal state in the following way,{h}_k^n = f_{LSTM}({h}_{k-1}^n, {\psi }_k),where {\psi }_k is computed by the attention-based weighted sum of the inputs {h}^p_j as in , i.e.{\psi }_k = \sum \limits _{j=1}^{\tau } softmax({{h}_{k-1}^n}^T {h}_j^p) {h}_j^pThen, t...
{ "cite_spans": [ { "arxiv_id": "", "doi": "", "end": 389, "openalex_id": "https://openalex.org/W2964308564", "raw": "D. Bahdanau, K. Cho, and Y. Bengio, “Neural machine translation by jointly learning to align and translate,” arXiv preprint arXiv:1409.0473, 2014.", "source_ref...
10.1109/TSE.2020.3020502
1810.00314
CODIT: Code Editing with Tree-Based Neural Models
[ "Saikat Chakraborty", "Yangruibo Ding", "Miltiadis Allamanis", "Baishakhi Ray" ]
[ "cs.SE" ]
2,018
en
Computer Science
[ -0.01377343013882637, -0.026600651443004608, -0.021564383059740067, -0.020908143371343613, -0.008355625905096531, 0.024662451818585396, 0.031560610979795456, -0.004376211669296026, 0.01217098068445921, 0.03702419996261597, -0.03153008967638016, 0.029485059902071953, 0.014002351090312004, 0...
c10cca8a275efa8ffe78e0229f6b53aab97bb9cb
subsection
10
29
Token Generation Model (
We now focus on generating a concrete code fragment c, i.e. a sequence of tokens (x_1, x_2, ...). For the edit task, the probability of an edited token x^n_k depends not only on the tokens of the previous version (x_1^p, ..., x_m^p) but also on the previously generated tokens x_1^n, ..., x_{k-1}^n. The next token x^n_k...
{ "cite_spans": [] }
10.1109/TSE.2020.3020502
1810.00314
CODIT: Code Editing with Tree-Based Neural Models
[ "Saikat Chakraborty", "Yangruibo Ding", "Miltiadis Allamanis", "Baishakhi Ray" ]
[ "cs.SE" ]
2,018
en
Computer Science
[ -0.01940368115901947, -0.003369330195710063, -0.028678396716713905, 0.0007565261912532151, -0.009648449718952179, -0.009503532201051712, 0.010273883119225502, 0.03368186205625534, 0.019007064402103424, 0.013301895000040531, -0.04716680943965912, 0.008946744725108147, 0.010327273979783058, ...
e59f47356f0a339bdcd21b1911d06e4bc1a70381
subsection
11
29
Token Generation Model (
Since the language grammar provides this information, we create a mask (mask(\theta _k^n)) that returns a -\infty value for masked entries and zero otherwise. Similarly, not all variable, method names, type names are valid at every position. We refine the mask based on the variables, method names and type names extract...
{ "cite_spans": [] }
10.1109/TSE.2020.3020502
1810.00314
CODIT: Code Editing with Tree-Based Neural Models
[ "Saikat Chakraborty", "Yangruibo Ding", "Miltiadis Allamanis", "Baishakhi Ray" ]
[ "cs.SE" ]
2,018
en
Computer Science
[ -0.027718355879187584, -0.02391776628792286, -0.02175036072731018, 0.03623533993959427, -0.012523628771305084, -0.04420284181833267, 0.030832091346383095, -0.013042584992945194, 0.00422033341601491, 0.008654354140162468, -0.020926136523485184, 0.003794865682721138, -0.030129974707961082, -...
83f87bb59a1ac4cd5d48123aafd84bc68790649b
subsection
12
29
Implementation
Our tree-based translation model is implemented as an edit recommendation tool, Codit. Codit learns source code changes from a dataset of patches. Then, given a code fragment to edit, Codit predicts potential changes that are likely to take place in a similar context. We implement Codit extending OpenNMT  based on PyTo...
{ "cite_spans": [ { "arxiv_id": "", "doi": "10.18653/v1/p17-4012", "end": 324, "openalex_id": "https://openalex.org/W2963212250", "raw": "G. Klein, Y. Kim, Y. Deng, J. Senellart, and A. M. Rush, “OpenNMT: Open-Source Toolkit for Neural Machine Translation,” ArXiv e-prints.", "s...
10.1109/TSE.2020.3020502
1810.00314
CODIT: Code Editing with Tree-Based Neural Models
[ "Saikat Chakraborty", "Yangruibo Ding", "Miltiadis Allamanis", "Baishakhi Ray" ]
[ "cs.SE" ]
2,018
en
Computer Science
[ -0.02783014439046383, -0.0231917854398489, -0.04363717511296272, 0.03893778845667839, 0.0017260333988815546, -0.07177247852087021, 0.021223535761237144, 0.002933303127065301, 0.005652998108416796, 0.03683221712708473, -0.030530765652656555, -0.0304087046533823, -0.023832611739635468, 0.024...
6d10dd7a13a62ad685d2ae039c43059ff46459c6
subsection
13
29
Implementation
Note that the losses of the two models are independent and thus we train each model separately. In our preliminary experiment, we found that the quality of the generated code is not entirely correlated to the loss. To mitigate this, we used top-1 accuracy to validate our model. We train the model for a fixed amount of ...
{ "cite_spans": [ { "arxiv_id": "", "doi": "", "end": 642, "openalex_id": "", "raw": "D. R. Reddy et al., “Speech understanding systems: A summary of results of the five-year research effort,” Department of Computer Science. Camegie-Mell University, Pittsburgh, PA, 1977.", "sou...
10.1109/TSE.2020.3020502
1810.00314
CODIT: Code Editing with Tree-Based Neural Models
[ "Saikat Chakraborty", "Yangruibo Ding", "Miltiadis Allamanis", "Baishakhi Ray" ]
[ "cs.SE" ]
2,018
en
Computer Science
[ -0.020914562046527863, -0.012631113640964031, -0.046863261610269547, 0.01809239201247692, -0.0037183985114097595, -0.040364645421504974, 0.038839150220155716, 0.0011193333193659782, -0.01261585857719183, 0.04503266513347626, -0.047717541456222534, -0.027992866933345795, -0.009450453333556652...
56049c67944ae2534cdf6795ec32ab12e48b49b6
subsection
14
29
Experimental Design
We evaluate Codit for three different types of changes that often appear in practice: (i) code change in the wild, (ii) pull request edits, and (iii) bug repair. For each task, we train and evaluate Codit on different datasets. tab:datasetsummary provides detailed statistics of the datasets we used.(i) Code Change Task...
{ "cite_spans": [ { "arxiv_id": "", "doi": "10.1109/msr.2017.24", "end": 524, "openalex_id": "https://openalex.org/W2703757306", "raw": "M. Beller, G. Gousios, and A. Zaidman, “TravisTorrent: Synthesizing Travis CI and GitHub for full-stack research on continuous integration,” in Pro...
10.1109/TSE.2020.3020502
1810.00314
CODIT: Code Editing with Tree-Based Neural Models
[ "Saikat Chakraborty", "Yangruibo Ding", "Miltiadis Allamanis", "Baishakhi Ray" ]
[ "cs.SE" ]
2,018
en
Computer Science
[ -0.05880197510123253, -0.016249118372797966, -0.04638246074318886, 0.033108532428741455, 0.0008939875988289714, -0.033901918679475784, 0.005137925501912832, -0.016828900203108788, -0.014052054844796658, 0.010115648619830608, -0.0029790052212774754, -0.021314572542905807, -0.02598333358764648...
520804c6a9d8d82871edcec0e931358cd7903f4b
subsection
15
29
Evaluation Metric
To evaluate Codit, we measure for a given code fragment, how accurately Codit generates patches. We consider Codit to correctly generate a patch if it exactly matches the developer produced patches. Codit produces the top K patches and we compute Codit's accuracy by counting how many patches are correctly generated in ...
{ "cite_spans": [] }
10.1109/TSE.2020.3020502
1810.00314
CODIT: Code Editing with Tree-Based Neural Models
[ "Saikat Chakraborty", "Yangruibo Ding", "Miltiadis Allamanis", "Baishakhi Ray" ]
[ "cs.SE" ]
2,018
en
Computer Science
[ -0.0314258448779583, -0.0007470312411896884, -0.0154688386246562, 0.032341159880161285, -0.00731489947065711, -0.06907583773136139, 0.00927520077675581, -0.011044811457395554, -0.012570338323712349, 0.05491895228624344, -0.03865684196352959, -0.01885550655424595, -0.015209498815238476, 0.0...
05de6c24faeb72cae75bab6d9214b8ee34bd2378
subsection
16
29
Baseline
We consider several baselines to evaluate Codit's performance. Our first baseline in a vanilla LSTM based Seq2Seq model with attention mechanism . Results of this baseline indicate different drawbacks of considering raw code as a sequence of token.The second baseline, we consider, is proposed by Tufano et al.  . For a ...
{ "cite_spans": [ { "arxiv_id": "", "doi": "", "end": 146, "openalex_id": "https://openalex.org/W2964308564", "raw": "D. Bahdanau, K. Cho, and Y. Bengio, “Neural machine translation by jointly learning to align and translate,” arXiv preprint arXiv:1409.0473, 2014.", "source_ref...
10.1109/TSE.2020.3020502
1810.00314
CODIT: Code Editing with Tree-Based Neural Models
[ "Saikat Chakraborty", "Yangruibo Ding", "Miltiadis Allamanis", "Baishakhi Ray" ]
[ "cs.SE" ]
2,018
en
Computer Science
[ -0.04134276881814003, 0.018352527171373367, -0.0550423264503479, 0.05565255135297775, -0.004889430478215218, -0.04253270849585533, 0.02094598487019539, -0.001283380202949047, -0.0013014962896704674, 0.022563083097338676, -0.02951965108513832, -0.0000305112680507591, -0.025827787816524506, ...
4073c9cd8c94edc0998388084a7d6797277401b0
subsection
17
29
Results
We evaluate Codit's performances to generate concrete patches w.r.t. generic edits (RQ1) and bug fixes (RQ3). In RQ2, we present an ablation study to evaluate our design choices.RQ1.  How accurately can Codit suggest concrete edits? [Table: Performance of Codit suggesting concrete patches. For Token Based models, pred...
{ "cite_spans": [ { "arxiv_id": "", "doi": "10.1145/3238147.3240732", "end": 1198, "openalex_id": "https://openalex.org/W2888328667", "raw": "M. Tufano, C. Watson, G. Bavota, M. Di Penta, M. White, and D. Poshyvanyk, “An empirical investigation into learning bug-fixing patches in the...
10.1109/TSE.2020.3020502
1810.00314
CODIT: Code Editing with Tree-Based Neural Models
[ "Saikat Chakraborty", "Yangruibo Ding", "Miltiadis Allamanis", "Baishakhi Ray" ]
[ "cs.SE" ]
2,018
en
Computer Science
[ -0.04779772832989693, 0.026584817096590996, -0.038824208080768585, 0.0192900151014328, 0.011773929931223392, -0.06482910364866257, -0.015047433786094189, -0.0012371019693091512, -0.012689595110714436, 0.028553497046232224, -0.018420133739709854, -0.017153464257717133, -0.03403222933411598, ...
a003903f051354ef9ec6ecbfcc322aca383ff168
subsection
18
29
Results
Instead, we present the context to the Codit through the token mask (see eqn:maskedprob). If we enable copy attention, Codit becomes highly biased by the tokens that are inside c_p. [Figure: Patch size (Tree Edit-distance) histogram of correctly generated patches in different datasets.][Table: Examples of different typ...
{ "cite_spans": [ { "arxiv_id": "", "doi": "10.1016/0022-0000(80)90002-1", "end": 1555, "openalex_id": "https://openalex.org/W2012010763", "raw": "W. J. Masek and M. S. Paterson, “A faster algorithm computing string edit distances,” Journal of Computer and System sciences, vol. 20, n...
10.1109/TSE.2020.3020502
1810.00314
CODIT: Code Editing with Tree-Based Neural Models
[ "Saikat Chakraborty", "Yangruibo Ding", "Miltiadis Allamanis", "Baishakhi Ray" ]
[ "cs.SE" ]
2,018
en
Computer Science
[ -0.026063410565257072, 0.01101743709295988, -0.0406821146607399, 0.01334452722221613, 0.008392784744501114, -0.05758976191282272, -0.011353149078786373, -0.008011294528841972, 0.010758023709058762, 0.01719757914543152, -0.039369791746139526, -0.012016941793262959, -0.025056276470422745, -0...
0553efb73b2bd9321dc59dccc0d2455d6896652b
subsection
19
29
Results
Other structural transformation that Codit include, but not limited to, include scoping (example 7 in examples), adding/deleting method parameters (example 3 in examples), changing method/variable access modifiers (example 9, 10 in examples), etc.Result : Codit suggests 15.94% correct patches for Code-Change-Data and 2...
{ "cite_spans": [ { "arxiv_id": "", "doi": "10.1145/3238147.3240732", "end": 2683, "openalex_id": "https://openalex.org/W2888328667", "raw": "M. Tufano, C. Watson, G. Bavota, M. Di Penta, M. White, and D. Poshyvanyk, “An empirical investigation into learning bug-fixing patches in the...
10.1109/TSE.2020.3020502
1810.00314
CODIT: Code Editing with Tree-Based Neural Models
[ "Saikat Chakraborty", "Yangruibo Ding", "Miltiadis Allamanis", "Baishakhi Ray" ]
[ "cs.SE" ]
2,018
en
Computer Science
[ -0.036769166588783264, 0.013937192969024181, -0.046075884252786636, 0.03530450165271759, 0.0023324009962379932, -0.07274497300386429, -0.023571934551000595, -0.008475217036902905, 0.0032840510830283165, 0.019269485026597977, -0.02601304091513157, -0.01096209418028593, -0.030315492302179337, ...
fdd349aee06f140d2f048f82c136cdc145d0a0c3
subsection
20
29
Results
For example, 3050 test patches of Code-Change-Data, and 225 test patches of Pull-Request-Data do not have structural changes. When we use these patches to train {M}_{tree}, we essentially train the model to sometimes copy the input to the output and rewarding the loss function for predicting no transformation. Thus, to...
{ "cite_spans": [] }
10.1109/TSE.2020.3020502
1810.00314
CODIT: Code Editing with Tree-Based Neural Models
[ "Saikat Chakraborty", "Yangruibo Ding", "Miltiadis Allamanis", "Baishakhi Ray" ]
[ "cs.SE" ]
2,018
en
Computer Science
[ -0.037454504519701004, -0.003283374011516571, -0.02614489197731018, 0.0005003281985409558, 0.007520662620663643, -0.05723487585783005, 0.01596471481025219, 0.016010502353310585, -0.0225886981934309, -0.010470166802406311, -0.044658467173576355, -0.015209214761853218, -0.025458073243498802, ...
60e4e4551e1a055c677ecae1cc2fa2c23702db3d
subsection
21
29
Results
With this, Codit generates 820 (15.94%), and 177 (28.87%) correct patches from Code-Change-Data and Pull-Request-Data respectively (tab:replaceunksummary).Second, we test two configuration parameters related to the beam size, K_{tree} and K_{token} i.e. the number of trees generated by {M}_{tree} and number of concrete...
{ "cite_spans": [] }
10.1109/TSE.2020.3020502
1810.00314
CODIT: Code Editing with Tree-Based Neural Models
[ "Saikat Chakraborty", "Yangruibo Ding", "Miltiadis Allamanis", "Baishakhi Ray" ]
[ "cs.SE" ]
2,018
en
Computer Science
[ -0.03021661564707756, 0.0015928582288324833, -0.08723139762878418, 0.023715464398264885, -0.010964968241751194, -0.053230077028274536, -0.028965219855308533, -0.022280938923358917, -0.005154374521225691, 0.01771792396903038, -0.019152451306581497, 0.007195522077381611, -0.016558093950152397,...
e00fc238583edabfa535405e84de6cf004e025a0
subsection
22
29
Results
The main reason behind such deterioration is diverse choice of token name across different projects. Developer tend to use project specific naming convention, api etc. This also indicates that the structural change pattern that developers follow are more ubiquitous across different projects than the token changes.Resul...
{ "cite_spans": [ { "arxiv_id": "", "doi": "10.1145/2610384.2628055", "end": 813, "openalex_id": "https://openalex.org/W2156723666", "raw": "R. Just, D. Jalali, and M. D. Ernst, “Defects4J: A database of existing faults to enable controlled testing studies for java programs,” in Proc...
10.1109/TSE.2020.3020502
1810.00314
CODIT: Code Editing with Tree-Based Neural Models
[ "Saikat Chakraborty", "Yangruibo Ding", "Miltiadis Allamanis", "Baishakhi Ray" ]
[ "cs.SE" ]
2,018
en
Computer Science
[ -0.03393254056572914, -0.001145261456258595, -0.03197959065437317, -0.000034269636671524495, 0.006717086769640446, -0.056208401918411255, -0.008834056556224823, -0.03805204853415489, -0.02380160242319107, 0.023145532235503197, -0.009238378144800663, -0.001930065918713808, -0.0264869108796119...
3a732497f89ce21d37ac208e2a7887c09ac043d3
subsection
23
29
Results
We see that, 48 out of 51 successful patches are generated within 20 minutes.We further manually compare the patches with the developer-provided patches: among 51 potential patches, 30 patches are identical and come from 25 different bug ids (See Table REF ). The bugs marked in [rgb]0.0,0.6,0green are completely fixed ...
{ "cite_spans": [ { "arxiv_id": "", "doi": "10.1109/tse.2019.2940179", "end": 1095, "openalex_id": "https://openalex.org/W2907705732", "raw": "Z. Chen, S. Kommrusch, M. Tufano, L.-N. Pouchet, D. Poshyvanyk, and M. Monperrus, “Sequencer: Sequence-to-sequence learning for end-to-end pr...
10.1109/TSE.2020.3020502
1810.00314
CODIT: Code Editing with Tree-Based Neural Models
[ "Saikat Chakraborty", "Yangruibo Ding", "Miltiadis Allamanis", "Baishakhi Ray" ]
[ "cs.SE" ]
2,018
en
Computer Science
[ -0.046489596366882324, 0.011050054803490639, -0.018940770998597145, 0.017551884055137634, -0.008226491510868073, -0.041117195039987564, 0.004872555378824472, -0.015445657074451447, -0.020161772146821022, 0.01541513204574585, -0.027716713026165962, 0.005998165346682072, -0.014690162613987923,...
4b8c26c25c977a19ac01a86a66d8d8a4d21b03d1
subsection
24
29
Threats to validity
External Validity. We built and trained Codit on real-world changes. Like all machine learning models, our hypothesis is that the dataset is representative of real code changes. To mitigate this threat, we collected patch data from different repositories and different types of edits collected from real world.Most NMT b...
{ "cite_spans": [ { "arxiv_id": "", "doi": "10.1109/tse.2019.2940179", "end": 786, "openalex_id": "https://openalex.org/W2907705732", "raw": "Z. Chen, S. Kommrusch, M. Tufano, L.-N. Pouchet, D. Poshyvanyk, and M. Monperrus, “Sequencer: Sequence-to-sequence learning for end-to-end pro...
10.1109/TSE.2020.3020502
1810.00314
CODIT: Code Editing with Tree-Based Neural Models
[ "Saikat Chakraborty", "Yangruibo Ding", "Miltiadis Allamanis", "Baishakhi Ray" ]
[ "cs.SE" ]
2,018
en
Computer Science
[ -0.016416113823652267, 0.016965351998806, -0.01537103671580553, 0.06316237151622772, -0.035700473934412, -0.04348134249448776, 0.0039819758385419846, -0.000803356000687927, 0.005969149526208639, 0.012182404287159443, -0.04232184216380119, -0.02376980148255825, -0.029658854007720947, -0.011...
8464291fa1842778545d89ec733599447c68f0dc
subsection
25
29
Related Work
Modeling source code. Applying ML to source code has received increasing attention in recent years  across many applications such as code completion , , bug prediction , , , clone detection , code search , etc. In these work, code was represented in many form, e.g., token sequences , , parse-trees , , graphs , , embedd...
{ "cite_spans": [ { "arxiv_id": "", "doi": "10.1145/3212695", "end": 451, "openalex_id": "https://openalex.org/W2963935794", "raw": "M. Allamanis, E. T. Barr, P. Devanbu, and C. Sutton, “A survey of machine learning for big code and naturalness,” ACM Computing Surveys, 2018.", ...
10.1109/TSE.2020.3020502
1810.00314
CODIT: Code Editing with Tree-Based Neural Models
[ "Saikat Chakraborty", "Yangruibo Ding", "Miltiadis Allamanis", "Baishakhi Ray" ]
[ "cs.SE" ]
2,018
en
Computer Science
[ -0.06659197807312012, 0.03384532779455185, -0.03949130326509476, 0.014427753165364265, 0.0013695302186533809, -0.05673441290855408, 0.00027776858769357204, 0.00794251263141632, 0.009743120521306992, 0.008758890442550182, -0.024369245395064354, -0.024765990674495697, -0.005260674748569727, ...
5b4e0e8582b43e2248f35d59bab50fdebe51affe
subsection
26
29
Related Work
Automatic program repair is a well-researched field, and previous researchers proposed many generic techniques for general software bugs repair , , , , . There are two differnt directions in program repair research : generate and validate approach, and sysnthesis bases approach. In generate and validate approaches, can...
{ "cite_spans": [ { "arxiv_id": "", "doi": "10.1145/2568225.2568258", "end": 153, "openalex_id": "https://openalex.org/W2009526138", "raw": "S. Kaleeswaran, V. Tulsian, A. Kanade, and A. Orso, “Minthint: Automated synthesis of repair hints,” in Proceedings of the 36th International C...
10.1109/TSE.2020.3020502
1810.00314
CODIT: Code Editing with Tree-Based Neural Models
[ "Saikat Chakraborty", "Yangruibo Ding", "Miltiadis Allamanis", "Baishakhi Ray" ]
[ "cs.SE" ]
2,018
en
Computer Science
[ -0.04082684591412544, 0.005595383234322071, -0.024609006941318512, 0.026073645800352097, 0.003564339131116867, -0.03603624925017357, -0.009009062312543392, 0.003274462418630719, 0.012685920111835003, 0.023602066561579704, -0.02715686894953251, -0.026073645800352097, -0.028880873695015907, ...
02b0db3ba2c666a6ec3fa02ad9f4f5e471e3e318
subsection
27
29
Discussion and Future Work
Search Space for Code Generation. Synthesizing patches (or code in general) is challenging . When we view code generation as a sequence of token generation problem, the space of the possible actions becomes too large. Existing statistical language modeling techniques endorse the action space with a probability distribu...
{ "cite_spans": [ { "arxiv_id": "", "doi": "", "end": 92, "openalex_id": "https://openalex.org/W1776302608", "raw": "P. Flener, Logic program synthesis from incomplete information. Springer Science & Business Media, 2012, vol. 295.", "source_ref_id": "ae53f9bf4a6dde6b72e3c7f179...
10.1109/TSE.2020.3020502
1810.00314
CODIT: Code Editing with Tree-Based Neural Models
[ "Saikat Chakraborty", "Yangruibo Ding", "Miltiadis Allamanis", "Baishakhi Ray" ]
[ "cs.SE" ]
2,018
en
Computer Science
[ -0.04070990905165672, -0.006053844466805458, -0.02923545241355896, 0.017181169241666794, 0.00808705110102892, -0.05059747397899628, 0.003719280706718564, 0.019286854192614555, -0.02041598968207836, 0.0271602850407362, -0.015334879979491234, -0.03958077356219292, -0.015899447724223137, -0.0...
bad47ff748055190b4aa75fd3339e500eba3c4a3
subsection
28
29
Conclusion
In this paper, we proposed and evaluated Codit, a tree-based hierarchical model for suggesting eminent source code changes. Codit's objective is to suggest changes that are similar to change patterns observed in the wild. We evaluate our work against a large number of real-world patches. The results indicate that tree-...
{ "cite_spans": [] }
10.1109/TSE.2020.3020502
1810.00314
CODIT: Code Editing with Tree-Based Neural Models
[ "Saikat Chakraborty", "Yangruibo Ding", "Miltiadis Allamanis", "Baishakhi Ray" ]
[ "cs.SE" ]
2,018
en
Computer Science
[ -0.0732254683971405, 0.001346280798316002, -0.025918764993548393, 0.011975415982306004, 0.0043744589202106, -0.08268376439809799, 0.02164728008210659, 0.01771141029894352, 0.00013872794806957245, 0.031883589923381805, -0.022791428491473198, -0.017787687480449677, -0.01951153762638569, 0.00...
4ccdcc29d6a0d4aed693d6db3e8a09c40c0cc3f7
abstract
0
98
Abstract
The free closed semialgebraic set $D_f$ determined by a hermitian noncommutative polynomial $f$ is the closure of the connected component of $\{(X,X^*)\mid f(X,X^*)>0\}$ containing the origin. When $L$ is a hermitian monic linear pencil, the free closed semialgebraic set $D_L$ is the feasible set of the linear matrix i...
{ "cite_spans": [] }
10.1007/s10208-020-09465-w
1808.06669
Noncommutative polynomials describing convex sets
[ "J. W. Helton", "I. Klep", "S. McCullough", "J. Volčič" ]
[ "math.FA", "math.OC" ]
2,018
en
Mathematics
[ -0.03675045073032379, 0.009904918260872364, -0.004307647235691547, -0.037910349667072296, 0.020786592736840248, -0.030722035095095634, 0.05561405420303345, 0.024525737389922142, 0.08528302609920502, 0.011522670276463032, -0.009912549518048763, 0.006020786240696907, -0.012011048384010792, 0...
6d4be751357da9061e50f94405c60c0bd2b85d4d
subsection
1
98
Introduction
Semidefinite programming (SDP) , is the main branch of convex optimization to emerge in the last 25 years. Feasibility sets of semidefinite programs are given by linear matrix inequalities (LMIs) and are called spectrahedra. We refer to the book for an overview of the substantial theory of LMIs and spectrahedra and the...
{ "cite_spans": [] }
10.1007/s10208-020-09465-w
1808.06669
Noncommutative polynomials describing convex sets
[ "J. W. Helton", "I. Klep", "S. McCullough", "J. Volčič" ]
[ "math.FA", "math.OC" ]
2,018
en
Mathematics
[ -0.03027869388461113, 0.013742925599217415, -0.03265947848558426, 0.012956961989402771, -0.0080198859795928, -0.015017256140708923, 0.04895869642496109, 0.02188490331172943, 0.061320461332798004, 0.05063745379447937, -0.024998236447572708, 0.007581119891256094, -0.01297985389828682, 0.0228...
da6aaf6c8d2358385f595a10e08b86e0e8aca434
subsection
2
98
Definitions
Let x=(x_1,\dots ,x_g) denote freely noncommuting variables and x^*=(x_1^*,\dots ,x_g^*) their formal adjoints. Let \!\mathop {<}\!x,x^*\!\mathop {>} denote the set of words in x and x^* and \mathop {<}\!x,x^*\!\mathop {>} the free polynomials in (x,x^*) equal the finite -linear combinations from <x,x*>. For a positive...
{ "cite_spans": [] }
10.1007/s10208-020-09465-w
1808.06669
Noncommutative polynomials describing convex sets
[ "J. W. Helton", "I. Klep", "S. McCullough", "J. Volčič" ]
[ "math.FA", "math.OC" ]
2,018
en
Mathematics
[ -0.020909788087010384, -0.011439332738518715, -0.02652643248438835, -0.02759481593966484, -0.0014385018730536103, -0.033730387687683105, 0.008699692785739899, 0.01794883981347084, 0.039347030222415924, 0.021795019507408142, -0.02199343405663967, -0.03333355858922005, -0.02916686423122883, ...
6e362d7864a8196e4d4d8bc087302ad75a2c9fd0
subsection
3
98
Definitions
A monic (linear) pencil of size \delta is an element L of \operatorname{M}_{{\delta }}(\mathop {<}\!x,x^*\!\mathop {>}) of the formL(x,x^*) = I_\delta - A\operatorname{\raisebox {1pt}{{\bigodot }}}x - B\operatorname{\raisebox {1pt}{{\bigodot }}}x^*= I_\delta - \sum _{j=1}^g A_j x_j - \sum _{j=1}^g B_j x_j^*.In the case...
{ "cite_spans": [] }
10.1007/s10208-020-09465-w
1808.06669
Noncommutative polynomials describing convex sets
[ "J. W. Helton", "I. Klep", "S. McCullough", "J. Volčič" ]
[ "math.FA", "math.OC" ]
2,018
en
Mathematics
[ -0.05760762840509415, 0.014333253726363182, -0.026896782219409943, -0.05010154843330383, 0.03170250356197357, -0.053183313459157944, 0.010145410895347595, 0.033624790608882904, 0.06987366080284119, 0.011358283460140228, -0.025905124843120575, 0.027095112949609756, 0.0041535161435604095, 0....
cabe3771319256b1ffb80749c648d2b82faf9d87
subsection
4
98
Main results
We are now ready to exposit our main results. Using the theory of realizations for noncommutative rational functions , , , , , in Theorem REF we explicitly and constructively describe the structure of noncommutative matrix polynomials f whose invertibility set \mathcal {K}_f is convex. Each \delta \times \delta noncomm...
{ "cite_spans": [] }
10.1007/s10208-020-09465-w
1808.06669
Noncommutative polynomials describing convex sets
[ "J. W. Helton", "I. Klep", "S. McCullough", "J. Volčič" ]
[ "math.FA", "math.OC" ]
2,018
en
Mathematics
[ -0.015644680708646774, 0.003960775677114725, -0.0026347932871431112, -0.016163626685738564, 0.0167130995541811, 0.0013851267285645008, 0.03544092923402786, 0.044781945645809174, 0.030709365382790565, 0.03244936093688011, -0.014034423977136612, 0.010356016457080841, -0.0009458354907110333, ...
7208c652286f2911b942f9dd2fed0336d19a21fb
subsection
5
98
Main results
The converse is proved in Section REF .Theorem REF implies that, for a monic linear pencil L, the invertibility set \mathcal {K}_L is convex if and only if the semisimple part of a minimal size pencil L describing \mathcal {K}_L is similar to a hermitian pencil.A non-invertible element f\in \operatorname{M}_{{\delta }...
{ "cite_spans": [] }
10.1007/s10208-020-09465-w
1808.06669
Noncommutative polynomials describing convex sets
[ "J. W. Helton", "I. Klep", "S. McCullough", "J. Volčič" ]
[ "math.FA", "math.OC" ]
2,018
en
Mathematics
[ -0.0679531842470169, 0.04087570309638977, 0.007643206976354122, -0.019445722922682762, 0.014897194691002369, -0.004464579280465841, 0.011020260863006115, 0.018713073804974556, 0.037487201392650604, 0.03782299906015396, -0.011401848867535591, 0.0014376327162608504, -0.004426420666277409, 0....
df2545d1961b6256849fe7bc647d292371c48e18
subsection
6
98
Main results
The proof of REF is based on Theorem REF (see Subsection REF ), while the proof of REF in Subsection REF uses REF and new, of independent interest, (recursive) certificates for invertibility of linear pencils on interiors of free spectrahedra.Theorem 1.4 (Nichtsingulärstellensatz) Let L be a hermitian monic pencil, an...
{ "cite_spans": [] }
10.1007/s10208-020-09465-w
1808.06669
Noncommutative polynomials describing convex sets
[ "J. W. Helton", "I. Klep", "S. McCullough", "J. Volčič" ]
[ "math.FA", "math.OC" ]
2,018
en
Mathematics
[ -0.03943672776222229, 0.025945214554667473, -0.011286168359220028, -0.0178716741502285, 0.037849489599466324, -0.03610963374376297, 0.04798338562250137, 0.012293453328311443, 0.06089494377374649, 0.023884858936071396, -0.012560536153614521, -0.016833866015076637, -0.009019777178764343, 0.0...
79d9d84a4b9c7dfe11fb8d2c7a856bf08f30931d
subsection
7
98
Main results
The matrix cube problem of , is a notable example of this phenomena , . See also , . Theorem REF provides another example as it gives a computationally tractable relaxation for the problem of determining whether a polynomial is of constant sign on the interior of a spectrahedron.
{ "cite_spans": [] }
10.1007/s10208-020-09465-w
1808.06669
Noncommutative polynomials describing convex sets
[ "J. W. Helton", "I. Klep", "S. McCullough", "J. Volčič" ]
[ "math.FA", "math.OC" ]
2,018
en
Mathematics
[ 0.01641506515443325, 0.014332671649754047, -0.017086314037442207, -0.008115999400615692, -0.005316589493304491, -0.025492170825600624, 0.051533542573451996, 0.031136756762862206, 0.023524193093180656, 0.010595040395855904, -0.0018487925408408046, -0.02215118519961834, -0.014202998019754887, ...
0d73df0a057858f337ce1b8812fdb5c01c8aaeba
subsection
8
98
Reader's guide
Section contains background and some preliminary results on linear pencils, free spectrahedra and realizations of noncommutative rational functions needed in the sequel. The proof of Theorem REF is given in Section , followed by the proof of Theorem REF and its corollary, Corollary REF , in Subsection REF . Corollary ...
{ "cite_spans": [] }
10.1007/s10208-020-09465-w
1808.06669
Noncommutative polynomials describing convex sets
[ "J. W. Helton", "I. Klep", "S. McCullough", "J. Volčič" ]
[ "math.FA", "math.OC" ]
2,018
en
Mathematics
[ -0.058242861181497574, 0.02007058821618557, 0.008356385864317417, -0.03849279507994652, 0.025061525404453278, -0.03574548661708832, 0.015797002241015434, 0.004716206807643175, 0.033181335777044296, 0.031059807166457176, -0.005536582320928574, 0.0243899617344141, -0.023809975013136864, 0.03...
c0efec0d671f64f652924a9502f6b035a45dbda2
subsection
9
98
Preliminaries
Let z=(z_1,\dots ,z_g,z_{g+1},\dots ,z_{2g})=(x_1,\dots ,x_g,y_1,\dots ,y_g) denote 2g freely noncommuting variables. Replacing z_{g+j}=y_j with x_j^* identifies \mathop {<}\!z \!\mathop {>} with \mathop {<}\!x,x^*\!\mathop {>}. On the other hand, elements f\in \mathop {<}\!z \!\mathop {>} are naturally evaluated at tu...
{ "cite_spans": [] }
10.1007/s10208-020-09465-w
1808.06669
Noncommutative polynomials describing convex sets
[ "J. W. Helton", "I. Klep", "S. McCullough", "J. Volčič" ]
[ "math.FA", "math.OC" ]
2,018
en
Mathematics
[ 0.002660695230588317, -0.03558274358510971, -0.026641283184289932, 0.0075681996531784534, 0.012862888164818287, -0.06164420768618584, 0.0346062034368515, 0.02769411727786064, -0.0018987255170941353, 0.02186538279056549, -0.0285028163343668, -0.029555650427937508, -0.03219536691904068, -0.0...
b339c60c0ef7f76c7e3b04e8b17d7a154c1811fe
subsection
10
98
Preliminaries
Therefore the results and definitions for matrix polynomials in z=(z_1,\dots ,z_{h}), whose assumptions refer only to the structure, and not to evaluations, of polynomials, directly apply to matrix polynomials in x_1,\dots ,x_g,x_1^*,\dots ,x_g^*.The free locus \mathcal {Z}_f of f\in \mathop {<}\!z \!\mathop {>}^{\del...
{ "cite_spans": [] }
10.1007/s10208-020-09465-w
1808.06669
Noncommutative polynomials describing convex sets
[ "J. W. Helton", "I. Klep", "S. McCullough", "J. Volčič" ]
[ "math.FA", "math.OC" ]
2,018
en
Mathematics
[ -0.01806071028113365, 0.001872425782494247, -0.038958657532930374, -0.017938677221536636, 0.024314839392900467, -0.03215538337826729, 0.01766410656273365, -0.0019439287716522813, 0.03499262407422066, -0.005113895982503891, -0.012881984002888203, 0.00013132720778230578, 0.007440127432346344, ...
5473d2f3dfa344a54513d216a80576a642f799d4
subsection
11
98
Preliminaries
If L and M are minimal and \mathcal {D}_L=\mathcal {D}_M, then L and M are unitarily equivalent. (See Proposition REF .) It is convenient to declare that the minimal pencil for the largest free spectrahedron \mathcal {D}_I=\lbrace (X,X^*)\colon X\in \operatorname{M}_{n}(^n,n\in {\mathbb {N}}\rbrace is of size 0. Every ...
{ "cite_spans": [] }
10.1007/s10208-020-09465-w
1808.06669
Noncommutative polynomials describing convex sets
[ "J. W. Helton", "I. Klep", "S. McCullough", "J. Volčič" ]
[ "math.FA", "math.OC" ]
2,018
en
Mathematics
[ -0.02727684937417507, 0.03670475259423256, -0.014340178109705448, -0.03618606552481651, 0.00428298395127058, -0.04967193678021431, 0.010564439930021763, 0.00681158434599638, 0.0439358651638031, 0.016628503799438477, -0.023615526035428047, -0.007501896005123854, 0.013020575977861881, -0.022...
6780459795417864b34c4337fde5b49dfbe9c459
subsection
12
98
Free loci and spectrahedra
For h,n\in {\mathbb {N}}, let \Omega ^{(n)}=(\Omega ^{(n)}_1,\dots ,\Omega ^{(n)}_h) be an h-tuple of n\times n generic matrices, that is,\Omega ^{(n)}_j=(\omega _{j\imath \jmath })_{\imath \jmath },where \omega _{j\imath \jmath } for 1\le j\le h and 1\le \imath ,\jmath \le n are commuting indeterminates.Lemma 2.1 A l...
{ "cite_spans": [] }
10.1007/s10208-020-09465-w
1808.06669
Noncommutative polynomials describing convex sets
[ "J. W. Helton", "I. Klep", "S. McCullough", "J. Volčič" ]
[ "math.FA", "math.OC" ]
2,018
en
Mathematics
[ -0.07185816764831543, 0.05516045540571213, 0.011462507769465446, -0.037150125950574875, 0.025183938443660736, -0.0387069508433342, -0.021322401240468025, 0.020177677273750305, 0.038310110569000244, -0.007337683811783791, -0.03449436277151108, 0.0009491673554293811, 0.005906778387725353, 0....
63434f4fd596a2852f01ba05bff0aa143296588e
subsection
13
98
Free loci and spectrahedra
By an invariant subspace for L, we mean an invariant subspace for \lbrace A_1,\dots ,A_g,A_1^*,\dots ,A_g^*\rbrace . Since L is hermitian, any invariant subspace for L is in fact reducing. Hence L=\oplus _i L^i, where each L^i is a hermitian monic pencil with no nontrivial invariant (equivalently reducing) subspaces. T...
{ "cite_spans": [] }
10.1007/s10208-020-09465-w
1808.06669
Noncommutative polynomials describing convex sets
[ "J. W. Helton", "I. Klep", "S. McCullough", "J. Volčič" ]
[ "math.FA", "math.OC" ]
2,018
en
Mathematics
[ -0.04889026656746864, 0.049714259803295135, 0.0013456649612635374, -0.0212101973593235, 0.028183622285723686, -0.03888027369976044, 0.0213627889752388, 0.0076524559408426285, 0.03396683558821678, 0.003982634283602238, -0.028168363496661186, 0.008758743293583393, -0.009514070115983486, 0.01...
bd58472b7358d518aa6527a881f245e1dc621259
subsection
14
98
Free loci and spectrahedra
If L(X,X^*)\nsucc 0, then there exists t\in (0,1) such that \det L(\gamma (t))=0, contradicting \mathcal {Z}_f=\mathcal {Z}_L. Therefore L(X,X^*)\succ 0. A similar argument shows L(X,X^*)\succ 0 implies (X,X^*)\in \mathcal {O}. Taking closures obtains \mathcal {K}_f=\mathcal {D}_L.Taking up items REF and REF , suppose ...
{ "cite_spans": [] }
10.1007/s10208-020-09465-w
1808.06669
Noncommutative polynomials describing convex sets
[ "J. W. Helton", "I. Klep", "S. McCullough", "J. Volčič" ]
[ "math.FA", "math.OC" ]
2,018
en
Mathematics
[ -0.036162104457616806, 0.0355822928249836, -0.02076650969684124, -0.0430283285677433, 0.0006747017614543438, -0.00736211659386754, -0.004127362743020058, -0.01071893610060215, 0.0010556817287579179, 0.005328951869159937, 0.00232307193800807, -0.0010814301203936338, -0.014213080517947674, -...
5d56bf5f94a241aef75028e9322fc01083244c2f
subsection
15
98
Realization theory
Let \operatorname{M}_{{\delta }}(z \leavevmode \vtop { {\hfil )\cr >#\hfil \cr )\crcr }} denote the \delta \times \delta noncommutative (nc) rational functions in z_1,\dots ,z_h , , . Evaluations and the involution for polynomials naturally extend to \operatorname{M}_{{\delta }}(z \leavevmode \vtop { {\hfil )\cr >#\hf...
{ "cite_spans": [] }
10.1007/s10208-020-09465-w
1808.06669
Noncommutative polynomials describing convex sets
[ "J. W. Helton", "I. Klep", "S. McCullough", "J. Volčič" ]
[ "math.FA", "math.OC" ]
2,018
en
Mathematics
[ -0.044670071452856064, -0.003060769522562623, -0.03899477422237396, 0.0007709134370088577, -0.005038350820541382, 0.01678178831934929, 0.03572995588183403, 0.06950712203979492, 0.022792721167206764, 0.01752934232354164, -0.030405551195144653, -0.017697159200906754, 0.0060986545868217945, 0...
41ecbbedb45db0924f709e60e8cbfccc0e795b7a
subsection
16
98
Realization theory
Given a realization I+c^* (I-A\operatorname{\raisebox {1pt}{{\bigodot }}}z)^{-1}(b\operatorname{\raisebox {1pt}{{\bigodot }}}z) there is a linear algebra algorithm – an extension of the Kalman decomposition – that produces a minimal realization. In the classical (commutative) one-variable setting, if \mathbb {r}(\zet...
{ "cite_spans": [] }
10.1007/s10208-020-09465-w
1808.06669
Noncommutative polynomials describing convex sets
[ "J. W. Helton", "I. Klep", "S. McCullough", "J. Volčič" ]
[ "math.FA", "math.OC" ]
2,018
en
Mathematics
[ -0.020345428958535194, 0.0012878060806542635, -0.03794353827834129, -0.005582401063293219, 0.012126363813877106, 0.027732666581869125, 0.0648977980017662, 0.03803511708974838, 0.02812950126826763, -0.013545813038945198, -0.04041612520813942, -0.0018038825364783406, -0.024390093982219696, 0...
1111b8b11a6eb66ee5ccd3615829ff188fad2b7d
subsection
17
98
Realization theory
By Remark REFREF , N_j:=A_j-b_jc^*, the coefficients of L+{\bf b}c^*, are the coefficients in a minimal realization of the polynomial f. By Remark REFREF , the N_j are jointly nilpotent. Hence \det f(\Omega ^{(n)})=\det L(\Omega ^{(n)}) for all n.(2) If 0\ne v\in \ker A, thenN_jv=-(c^*v)b_j,and c^*v\in \lbrace 0\rbrac...
{ "cite_spans": [] }
10.1007/s10208-020-09465-w
1808.06669
Noncommutative polynomials describing convex sets
[ "J. W. Helton", "I. Klep", "S. McCullough", "J. Volčič" ]
[ "math.FA", "math.OC" ]
2,018
en
Mathematics
[ 0.00402841717004776, 0.030716681852936745, -0.006073144264519215, -0.01820417307317257, 0.024780869483947754, -0.011085777543485165, 0.04348859563469887, 0.02621523104608059, 0.028855064883828163, 0.02703922614455223, -0.01699870079755783, 0.0059434412978589535, -0.04147438704967499, 0.030...
838078328a01b2dce8f706543c74d324c8b771a8
subsection
18
98
Proof of Theorem
We start the proof of Theorem REF with a lemma.Lemma 3.1 Suppose \mathbb {r}\in x,x^* \leavevmode \vtop { {\hfil )\cr >#\hfil \cr \setminus \crcr }} is defined at the origin and r(0)=1. Assume that r is hermitian and r=1+c* L-1b is a minimal FM realization, where b=j bjxj+j bjxj*. If L is irreducible and monic hermit...
{ "cite_spans": [] }
10.1007/s10208-020-09465-w
1808.06669
Noncommutative polynomials describing convex sets
[ "J. W. Helton", "I. Klep", "S. McCullough", "J. Volčič" ]
[ "math.FA", "math.OC" ]
2,018
en
Mathematics
[ -0.029198169708251953, 0.022409671917557716, -0.014484671875834465, -0.019404426217079163, 0.012669320218265057, 0.005285875406116247, 0.04417863115668297, 0.040792010724544525, 0.02016717940568924, 0.013912606984376907, -0.028908323496580124, -0.03194407746195793, -0.02352328971028328, 0....
92eef1ab46b01762a186486874c0939475d90967
subsection
19
98
Proof of Theorem
Hence by (REF ),(),() and the fact that w(A,A^*) span \operatorname{M}_{d}(, there exist \lambda _{jk}^1,\lambda _{jk}^2,\lambda _{jk}^3\in such that \begin{alignat}{3} {b}_j &=\lambda _{jk}^1 A_jc,&\qquad \widehat{b}_k&=\overline{\lambda _{jk}^1} A_k^*c \\ {b}_j &=\lambda _{jk}^2 A_jc,&\qquad {b}_k&=\overline{\lambda...
{ "cite_spans": [] }
10.1007/s10208-020-09465-w
1808.06669
Noncommutative polynomials describing convex sets
[ "J. W. Helton", "I. Klep", "S. McCullough", "J. Volčič" ]
[ "math.FA", "math.OC" ]
2,018
en
Mathematics
[ 0.0003222548693884164, 0.04060792550444603, -0.024498997256159782, -0.02323286049067974, 0.033133137971162796, -0.01443091593682766, 0.03539083153009415, -0.013225797563791275, 0.026832962408661842, 0.017542868852615356, -0.03523828461766243, -0.002707703737542033, -0.0224243625998497, 0.0...
56120c3d4201a22a40b1a4c3d2a8cfcb4b31f2c7
subsection
20
98
Proof of Theorem
By Remark REFREF , f admits a minimal realizationf=1 - \varepsilon c^*\Big ( I-A(I-\varepsilon cc^*)\operatorname{\raisebox {1pt}{{\bigodot }}}x-A^*(I-\varepsilon cc^*)\operatorname{\raisebox {1pt}{{\bigodot }}}x^* \Big )^{-1}(A\operatorname{\raisebox {1pt}{{\bigodot }}}x+A^*\operatorname{\raisebox {1pt}{{\bigodot }}}x...
{ "cite_spans": [] }
10.1007/s10208-020-09465-w
1808.06669
Noncommutative polynomials describing convex sets
[ "J. W. Helton", "I. Klep", "S. McCullough", "J. Volčič" ]
[ "math.FA", "math.OC" ]
2,018
en
Mathematics
[ 0.0029938523657619953, 0.05092218890786171, -0.025598391890525818, -0.008504829369485378, 0.016140107065439224, -0.04064011573791504, 0.09165383875370026, 0.017833445221185684, 0.030251258984208107, 0.008939605206251144, -0.04475904628634453, -0.01862671971321106, -0.035422805696725845, 0....
bcb99c3fc53292e19234d3729d8a29bb3dae5b58
subsection
21
98
Proof of Theorem
It follows that AP,A^*P are jointly nilpotent of order at most two and\Big (I-A(I-cc^*)\operatorname{\raisebox {1pt}{{\bigodot }}}x - A^*(I-cc^*)\operatorname{\raisebox {1pt}{{\bigodot }}}x^*\Big )^{-1} =I+A(I-cc^*)\operatorname{\raisebox {1pt}{{\bigodot }}}x + A^*(I-cc^*)\operatorname{\raisebox {1pt}{{\bigodot }}}x^*....
{ "cite_spans": [] }
10.1007/s10208-020-09465-w
1808.06669
Noncommutative polynomials describing convex sets
[ "J. W. Helton", "I. Klep", "S. McCullough", "J. Volčič" ]
[ "math.FA", "math.OC" ]
2,018
en
Mathematics
[ -0.010445975698530674, 0.008324733003973961, -0.00274693313986063, -0.047705069184303284, 0.020342566072940826, -0.048620715737342834, 0.06140921637415886, 0.03284111246466637, 0.050146788358688354, 0.014261160977184772, -0.02424931526184082, 0.0021651173010468483, -0.0161000806838274, -0....
a2d4d0c5d7c57121fea0b2141fa3bee9c7b2d571
subsection
22
98
Proof of Theorem
If \mathcal {D}_f=\mathcal {D}_L, then L is irreducible and there exists b_j,c\in d such that f^{-1}=I+ c^* L^{-1}{\bf b} is a minimal FM realization.Write L=I-A\operatorname{\raisebox {1pt}{{\bigodot }}}x-A^*\operatorname{\raisebox {1pt}{{\bigodot }}}x^*. By Proposition REFREF , \mathcal {Z}_f=\mathcal {Z}_L. After a ...
{ "cite_spans": [] }
10.1007/s10208-020-09465-w
1808.06669
Noncommutative polynomials describing convex sets
[ "J. W. Helton", "I. Klep", "S. McCullough", "J. Volčič" ]
[ "math.FA", "math.OC" ]
2,018
en
Mathematics
[ -0.021681837737560272, 0.047452859580516815, -0.0005311744753271341, -0.02270413376390934, 0.046262722462415695, -0.0290973000228405, 0.06008661165833473, 0.029326172545552254, 0.015021652914583683, 0.027754582464694977, -0.03326277807354927, -0.005378499161452055, -0.020659541711211205, 0...
869e2cdcd9e42d7472d4bfdcb40fe0057b731aea
subsection
23
98
Proof of Theorem
If \mathcal {D}_f is proper and convex, then f has degree two and is concave.Further, normalizing f(0)=1, if L is a minimal hermitian monic pencil such that \mathcal {D}_f=\mathcal {D}_L, then L is irreducible, f is a Schur complement of L and there exist vectors c,b_1,\dots , b_{2g} such thatf^{-1} = 1+ c^* L^{-1}{\bf...
{ "cite_spans": [] }
10.1007/s10208-020-09465-w
1808.06669
Noncommutative polynomials describing convex sets
[ "J. W. Helton", "I. Klep", "S. McCullough", "J. Volčič" ]
[ "math.FA", "math.OC" ]
2,018
en
Mathematics
[ -0.07942263782024384, 0.03897874802350998, 0.015673913061618805, -0.04322154074907303, 0.03047790192067623, 0.004467904567718506, 0.02397635579109192, -0.003134401049464941, 0.017886880785226822, 0.03302662819623947, -0.027318701148033142, 0.03314872458577156, 0.007920896634459496, 0.01997...
70c9ac38f84361aacc9c18be36219e97142e3020
subsection
24
98
Proof of Theorem
Therefore the set of irreducible components of \mathcal {Z}_{L}(n) contains the set of irreducible components of \mathcal {Z}_{\widetilde{L}}(n). Since\mathcal {Z}_L=\mathcal {Z}_{L^1}\cup \cdots \cup \mathcal {Z}_{L^\ell }and the \mathcal {Z}_{L^i}(n) are irreducible hypersurfaces for all n large enough by Lemma REF ,...
{ "cite_spans": [] }
10.1007/s10208-020-09465-w
1808.06669
Noncommutative polynomials describing convex sets
[ "J. W. Helton", "I. Klep", "S. McCullough", "J. Volčič" ]
[ "math.FA", "math.OC" ]
2,018
en
Mathematics
[ -0.04513552784919739, 0.06610111892223358, 0.01509858202189207, -0.019363416358828545, 0.020996106788516045, -0.013992318883538246, 0.03482057899236679, -0.0028400439769029617, 0.010581976734101772, 0.024566667154431343, -0.014030465856194496, 0.0018482219893485308, -0.03105165623128414, 0...
91dde4b889de7e58cb02365e6e58cf5f755f5abc
subsection
25
98
Proof of Theorem
If it is not similar to L^{i_k} for any k, then (REF ) implies\bigcap _k\mathcal {K}_{L^{i_k}}\subseteq \mathcal {K}_{L^m}.Hence \mathcal {K}_f=\mathcal {D}_{\widehat{L}} holds by (REF ).Remark 4.1 Given a factorization of f into atomic factors f=f_1\cdots f_t with f_j(0)=I, one can use the proof of Theorem REF to ide...
{ "cite_spans": [] }
10.1007/s10208-020-09465-w
1808.06669
Noncommutative polynomials describing convex sets
[ "J. W. Helton", "I. Klep", "S. McCullough", "J. Volčič" ]
[ "math.FA", "math.OC" ]
2,018
en
Mathematics
[ -0.030024034902453423, 0.058919116854667664, -0.012403221800923347, -0.03014608472585678, 0.019741414114832878, 0.012311684899032116, 0.020672036334872246, -0.006682178471237421, -0.01236508134752512, 0.02889508381485939, -0.026194751262664795, 0.00584309222176671, -0.024638626724481583, 0...
fb9ab18a17ff0f0a4a3add35ab4141606ed08343
subsection
26
98
Proof of Theorem
By Remark REFREF , \mathcal {K}_f=\mathcal {K}_L. By assumption there exists a minimal hermitian monic pencil \widetilde{L} such that \mathcal {K}_L=\mathcal {D}_{\widetilde{L}}. By \partial \mathcal {K}_L(n) we denote the topological boundary of \mathcal {K}_L(n). Thus\mathcal {Z}_L(n)\supseteq \partial \mathcal {K}_L...
{ "cite_spans": [] }
10.1007/s10208-020-09465-w
1808.06669
Noncommutative polynomials describing convex sets
[ "J. W. Helton", "I. Klep", "S. McCullough", "J. Volčič" ]
[ "math.FA", "math.OC" ]
2,018
en
Mathematics
[ -0.038794293999671936, 0.0684928372502327, 0.006985872983932495, -0.02808084338903427, 0.010881327092647552, -0.01442195475101471, 0.03787861764431, -0.011453626677393913, 0.0138267632573843, 0.02925596572458744, -0.0028080842457711697, 0.014101467095315456, -0.021930528804659843, 0.043006...
7288fec9b6c632981f06224fcef3b35fcf5fc692
subsection
27
98
Proof of Theorem
Then \mathcal {K}_{L^{i_k}}=\mathcal {D}_{\widetilde{L}^k} is convex for every k and therefore\mathcal {K}_L=\bigcap _k\mathcal {K}_{L^{i_k}}=\bigcap _k\mathcal {D}_{\widetilde{L}^k}=\mathcal {D}_{\widetilde{L}^1\oplus \cdots \oplus \widetilde{L}^s}.Moreover, L^{i_k} is similar to \widetilde{L}^k by .Recall that \wideh...
{ "cite_spans": [] }
10.1007/s10208-020-09465-w
1808.06669
Noncommutative polynomials describing convex sets
[ "J. W. Helton", "I. Klep", "S. McCullough", "J. Volčič" ]
[ "math.FA", "math.OC" ]
2,018
en
Mathematics
[ -0.025219494476914406, 0.056999415159225464, -0.0076551008969545364, -0.04857766255736351, 0.01934562623500824, 0.009161709807813168, 0.03576194494962692, -0.015867074951529503, -0.002128324005752802, 0.0502864234149456, -0.01497455220669508, 0.001595289446413517, -0.0334429107606411, 0.05...
fa1284674ad17f8d28e5514a93b32ebdbaf3bff0
subsection
28
98
Proof of Theorem
As in Remark REF , for every i there exists j_i such that \mathcal {Z}_{L^i}=\mathcal {Z}_{f_{j_i}}, whence \mathcal {K}_{L^i}=\mathcal {K}_{f_{j_i}}. If some L^i is not similar to a hermitian monic pencil, then {L} is nontrivial and is invertible on \operatorname{int}\mathcal {K}_{\widehat{L}} by convexity of \mathcal...
{ "cite_spans": [] }
10.1007/s10208-020-09465-w
1808.06669
Noncommutative polynomials describing convex sets
[ "J. W. Helton", "I. Klep", "S. McCullough", "J. Volčič" ]
[ "math.FA", "math.OC" ]
2,018
en
Mathematics
[ -0.033106546849012375, 0.05696767196059227, 0.014775894582271576, 0.0026908605359494686, 0.002458199393004179, -0.04592198505997658, 0.002736629918217659, -0.009901451878249645, 0.013311273418366909, 0.009329333901405334, -0.024074716493487358, 0.008032533340156078, -0.032831933349370956, ...
95264ff7c2572df468c4bd9a82a456d4dd0379c5
subsection
29
98
Finding an LMI representation for a convex
The main result of states that for a hermitian matrix polynomial f\in \operatorname{M}_{{\delta }}(\mathop {<}\!x,x^*\!\mathop {>}) with f(0)\succ 0, the set \mathcal {K}_f(n) is convex for all n if and only if \mathcal {K}_f is a free spectrahedron. Actually, the version in does this for hermitian f with bounded \math...
{ "cite_spans": [] }
10.1007/s10208-020-09465-w
1808.06669
Noncommutative polynomials describing convex sets
[ "J. W. Helton", "I. Klep", "S. McCullough", "J. Volčič" ]
[ "math.FA", "math.OC" ]
2,018
en
Mathematics
[ 0.011280768550932407, 0.010556175373494625, -0.006582354661077261, -0.019891982898116112, 0.026359928771853447, -0.04960792139172554, 0.026070091873407364, 0.021569987758994102, 0.03490249812602997, 0.029258299618959427, -0.0224242452532053, 0.0028983717784285545, -0.02066996693611145, -0....
2f692c3df84bac1daea2d6b2b15e4c03f6442dbf
subsection
30
98
Algorithm
We next explain how the machinery developed in this paper produces an explicit minimal LMI representation for a convex \mathcal {K}_f. This efficient algorithm only involves linear algebra and semidefinite programming (SDP) , .Compute the minimal realization I+c^*L^{-1}{\bf b} for f^{-1}. To construct this realization...
{ "cite_spans": [] }
10.1007/s10208-020-09465-w
1808.06669
Noncommutative polynomials describing convex sets
[ "J. W. Helton", "I. Klep", "S. McCullough", "J. Volčič" ]
[ "math.FA", "math.OC" ]
2,018
en
Mathematics
[ -0.06003188341856003, 0.046755895018577576, 0.003603209974244237, 0.013695632107555866, 0.011376149021089077, 0.002174515277147293, 0.03467016667127609, 0.015145308338105679, 0.042971476912498474, 0.04812927171587944, -0.04474160820245743, 0.0016356551786884665, 0.015297906473279, 0.062107...
6372960b514c7c438360595fe1a3bdf106dbfacf
subsection
31
98
Algorithm
Solve the following feasibility SDP:\begin{split} \operatorname{tr}(\operatorname{Re}(D\widetilde{L})(0))&=1\\ \operatorname{Re}(D\widetilde{L})& = P_0 + \sum _k C_k^* L C_k \quad \text{ for some } C_k,P_0, \text{ with }P_0\succeq 0. \end{split}We note that (REF ) is a SDP. Indeed, the first equation is simply a linear...
{ "cite_spans": [] }
10.1007/s10208-020-09465-w
1808.06669
Noncommutative polynomials describing convex sets
[ "J. W. Helton", "I. Klep", "S. McCullough", "J. Volčič" ]
[ "math.FA", "math.OC" ]
2,018
en
Mathematics
[ -0.02736601047217846, 0.04740592837333679, -0.018620487302541733, -0.005139712244272232, -0.0012057528365403414, 0.056655123829841614, 0.05464044585824013, 0.039316702634096146, 0.05378573387861252, 0.025961844250559807, -0.0274575874209404, -0.0017332697752863169, 0.005731141660362482, 0....
5ca6b9fce1482c2d53e4f1a3a0a4a163fb3ddd33
subsection
32
98
Algorithm
One first computes and mods out the radical of \mathcal {A} (corresponding to the \star entries) using the algorithm in ; then the algorithm of is applied to find the irreducible blocks L^j. Alternatively, gives an algorithm for decomposing \mathcal {A} as a direct sum of minimal left ideals; after omitting the ideals ...
{ "cite_spans": [] }
10.1007/s10208-020-09465-w
1808.06669
Noncommutative polynomials describing convex sets
[ "J. W. Helton", "I. Klep", "S. McCullough", "J. Volčič" ]
[ "math.FA", "math.OC" ]
2,018
en
Mathematics
[ -0.05032031983137131, 0.039822932332754135, 0.005420343484729528, -0.008216341026127338, 0.015257828868925571, -0.026258723810315132, 0.03951777517795563, -0.005904779769480228, 0.042843982577323914, 0.03384186327457428, -0.04089098051190376, -0.008086648769676685, -0.004111984744668007, 0...
a9e858fe131a0f6f4bd79598e8de92a0cf32961b
subsection
33
98
Algorithm
If \varepsilon ^{\prime }=\dim V>0, then let \widetilde{L}^{\prime } be the \delta \times \varepsilon ^{\prime } affine linear pencil whose coefficients are the restrictions of coefficients of \widetilde{L} to V. Then \widetilde{L} is of full rank on the interior of \mathcal {D}_L if and only if \widetilde{L}^{\prime }...
{ "cite_spans": [] }
10.1007/s10208-020-09465-w
1808.06669
Noncommutative polynomials describing convex sets
[ "J. W. Helton", "I. Klep", "S. McCullough", "J. Volčič" ]
[ "math.FA", "math.OC" ]
2,018
en
Mathematics
[ -0.0331735722720623, 0.011848794296383858, 0.015068492852151394, -0.006519507151097059, 0.02905358001589775, -0.02252262830734253, 0.03473001345992088, 0.0166478231549263, 0.03900260105729103, 0.016479970887303352, -0.018524708226323128, -0.007347320672124624, 0.014954048208892345, 0.01232...
c93293300a826348a2827a4ebf385e731a1e46f9
subsection
34
98
Checking whether
As a side product of Theorem REF and the Algorithm in Subsection REF we obtain a procedure for checking whether \mathcal {K}_f is convex.Given f\in \operatorname{M}_{{\delta }}(\mathop {<}\!x,x^*\!\mathop {>}) with f(0)=I, we construct the realization of f^{-1} and identify its irreducible blocks L^i, choosing one fr...
{ "cite_spans": [] }
10.1007/s10208-020-09465-w
1808.06669
Noncommutative polynomials describing convex sets
[ "J. W. Helton", "I. Klep", "S. McCullough", "J. Volčič" ]
[ "math.FA", "math.OC" ]
2,018
en
Mathematics
[ -0.047156088054180145, 0.014078152365982533, -0.00027708016568794847, -0.0186640452593565, 0.024478435516357422, -0.027561131864786148, 0.047339219599962234, 0.011354087851941586, 0.054756004363298416, 0.02626395784318447, -0.023211782798171043, -0.008034848608076572, -0.015344805084168911, ...
cafbd7f672e7b3dcf63a8a84a448876cb6f64c10
subsection
35
98
Checking whether
Consequently \widetilde{L}(X,X^*) has full rank.Proposition 4.3 Let \delta \ge \varepsilon . If every solution of (REF ) satisfiesP_0=0,\quad C_k=0\quad \text{for all }k,then there exists X\in \operatorname{M}_{\max \lbrace d,\varepsilon \rbrace }(^g such that L(X,X^*)\succ 0 and \ker \widetilde{L}(X,X^*)\ne \lbrace 0...
{ "cite_spans": [] }
10.1007/s10208-020-09465-w
1808.06669
Noncommutative polynomials describing convex sets
[ "J. W. Helton", "I. Klep", "S. McCullough", "J. Volčič" ]
[ "math.FA", "math.OC" ]
2,018
en
Mathematics
[ -0.03118853271007538, 0.00004035178790218197, -0.032287150621414185, -0.023116745054721832, 0.002840002765879035, -0.014091297052800655, 0.04376160353422165, 0.001630760496482253, 0.037597134709358215, 0.007404224947094917, 0.010429237969219685, 0.004249514080584049, 0.00027322390815243125, ...
5b6a5f0c4e633688da3b9b5ef2563bde91d9fc2f
subsection
36
98
Checking whether
Using the standard argument involving Caratheodory's theorem on convex hulls it is easy to show that \mathcal {C}+\mathcal {S} is closed in \mathcal {V}^{\rm h}_2; see e.g. .Lemma 4.4 Keep the notation from above. If every solution of (REF ) satisfiesP_0=0,\quad C_k=0\quad \text{for all }k,then \mathcal {U}\cap (\math...
{ "cite_spans": [] }
10.1007/s10208-020-09465-w
1808.06669
Noncommutative polynomials describing convex sets
[ "J. W. Helton", "I. Klep", "S. McCullough", "J. Volčič" ]
[ "math.FA", "math.OC" ]
2,018
en
Mathematics
[ -0.04986436292529106, 0.029067199677228928, -0.025298381224274635, 0.0020884897094219923, 0.008155600167810917, -0.0027236195746809244, -0.010833444073796272, 0.005840142257511616, 0.04177742451429367, 0.030837170779705048, -0.01942390762269497, 0.003122244495898485, 0.021727923303842545, ...