{"id":"a6e504d4-9f1a-4bc4-b36f-bfe739994867","arxiv_id":"1908.05406","paper_version":2,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":7.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"The Douglas-Rachford algorithm converges weakly to a normal solution of minimizing a convex function over a linear subspace even when the original problem is infeasible.","lead":"This paper proves that the Douglas-Rachford algorithm, a standard method for minimizing sums of convex functions, still converges in a useful way when the problem has no exact solution, as long as the constraint is a linear subspace. The iterates converge to what the authors call a normal solution, even when the objective's domain and the subspace are disjoint.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Infinite-dimensional identification of the minimal displacement vector is assumed, not proved; the theorem is conditional on (28).","rationale":"The reader's weakest_assumption identifies exactly the same point: the identity v = P_{U-dom g}(0) is proved only in finite dimensions and then assumed in general Hilbert spaces. I agree this is the most load-bearing unproved condition in the paper, since all convergence conclusions flow through v ∈ U^⊥. However, this is an explicitly stated hypothesis of Theorem 5.1, not a hidden error; the proof is internally consistent and the paper clearly flags the finite-dimensional result. The main theorem is conditional and remains correct under its stated assumptions. The concern affects the breadth of infinite-dimensional applicability, not the validity of the central argument. No fatal flaw or circularity was found. Therefore the reader's ACCEPT verdict should stand.","tokens_in":17637,"tokens_out":18183,"duration_ms":176458,"concrete_test":"Independently re-derive Proposition 3.1(ii) without finite-dimensionality, using [9, Proposition 6.1(ii) and Corollary 6.5(i)] under the standing assumptions (9) and (11). Specifically, attempt to construct an infinite-dimensional counterexample in ℓ2, e.g. U = span{e1} with g chosen so that 0 ∈ U^⊥ + dom g* and Z ≠ ∅, and compute both v = P_ran(Id-T)(0) and P_{U-dom g}(0). If an example with v ≠ P_{U-dom g}(0) exists, then (28) is a substantive restriction and Theorem 5.1 does not cover all infinite-dimensional instances satisfying (11); if no such example exists, the proof of the identity likely generalizes and (28) is redundant.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing soft spot is the assumption (28), introduced after Proposition 3.1: the paper assumes v = P_{U-dom g}(0) in general Hilbert space, while Proposition 3.1 proves this identity only in finite dimensions. Every subsequent result relies on the consequence (29), v ∈ U^⊥. Lemma 4.1(ii) needs (29) to conclude P_U T^n x - P_U T^{n+1}x → 0; Proposition 3.9 needs it to identify Z = P_U(F); Theorem 5.1 needs the whole chain to identify the weak limit as a normal solution. If (28) fails, v may have a nonzero U-component, and the shadow sequence can fail to converge to a minimizer of the shifted objective, as Example 5.5 demonstrates when (11) fails. The paper does not establish whether assumption (11) alone prevents such a failure in infinite dimensions, so the main theorem's scope in infinite-dimensional Hilbert spaces is narrower than the abstract suggests. The theorem itself is internally consistent and clearly conditional, but the infinite-dimensional applicability of the central claim rests on an unproved identity.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the Douglas-Rachford algorithm for minimizing the sum of an indicator function of a closed linear subspace U and a proper, lower semicontinuous, convex function g in a real Hilbert space, without assuming the sum has a minimizer. The authors introduce the minimal displacement vector v and the normal solution set Z, and under assumptions (9), (10), (11), and (28) — where (28) is proved only in finite dimensions — they establish in Theorem 5.1 that the shadow sequence P_U T^n x converges weakly to P_U y(x) in argmin(ι_U + g(·−v)), and that g(P_g R_U T^n x) converges to the corresponding infimum. The proof combines a function-value analysis (Lemma 4.2), a cluster-point argument (Lemma 4.3), and a uniqueness step using weak-to-weak continuity of P_Z. The paper also provides examples, including counterexamples to natural conjectures, and a parallel-splitting application in Section 6.","tokens_in":17771,"tokens_out":19754,"duration_ms":174533,"significance":"If the result holds as stated, it is a significant extension of Douglas-Rachford convergence theory to inconsistent convex optimization, going beyond prior work restricted to two indicator functions or affine subspaces. The paper contains detailed proofs, a careful statement of assumptions, and instructive examples that delineate the boundary of the theory. The identification of the shifted objective ι_U + g(·−v) is a conceptual contribution, and the parallel-splitting result in Section 6 is a useful application. The proof is mostly self-contained modulo cited facts, and the main theorem is precise and falsifiable.","major_comments":[{"comment":"Assumption (28), v = P_{U−dom g}(0), is stated without proof for infinite-dimensional Hilbert spaces; Proposition 3.1(ii) verifies it only when X is finite-dimensional. Since the derived property v ∈ U⊥ (29) is used in Proposition 3.2, Proposition 3.9, Lemma 4.1, Lemma 4.2, Lemma 4.3, and Theorem 5.1, the main theorem's applicability in infinite-dimensional Hilbert spaces is conditional on an unproved identity. The authors should either prove (28) under the standing assumptions (in particular (11)) or, if this is an open question, state it as an explicit limitation and adjust the abstract and introduction so that they do not suggest the result holds for every Hilbert space satisfying (11) alone.","section":"Section 3, after Proposition 3.1"},{"comment":"The proof of weak convergence in Theorem 5.1 relies on assumption (10), the weak-to-weak continuity of P_Z. This is a nontrivial restriction: for a general closed convex set Z in an infinite-dimensional Hilbert space, the metric projection need not be weak-to-weak continuous (for example, projection onto the unit ball in ℓ2). The paper does not provide sufficient conditions for (10) beyond the finite-dimensional case, and the infinite-dimensional examples do not verify it. Please add a discussion of (10) and, if possible, examples of infinite-dimensional settings where it holds.","section":"Theorem 5.1 and assumption (10)"}],"minor_comments":[{"comment":"The claim that all weak cluster points of (P_U T^n x) lie in U∩(v+dom g) is not justified as written, because dom g need not be weakly closed for a proper lower semicontinuous convex function (e.g., g(x)=1/x on R_{++}). The proof appears to rely on this property. Since this item is not used in the subsequent arguments (Lemma 4.3 uses only (i) and (v) together with weak lower semicontinuity of g), please correct the statement or remove it.","section":"Lemma 4.1(iv)"},{"comment":"The sentence \"Under the above assumptions, which we assume for the rest of the paper\" appears in the introduction before assumption (28) is introduced later in Section 3. Consider reordering the presentation so that all standing assumptions are listed before the statement of the main result.","section":"Introduction"},{"comment":"There are minor typographical issues, such as \"optimiz ation\" in the abstract and the unnecessary double spacing in \"the Douglas-Rachford algorithm\" in the introduction.","section":"Abstract and Section 5"},{"comment":"Example 5.2 claims a consequence of Theorem 5.1 without verifying that the example satisfies assumption (10). Please add a note explaining how (10) is obtained in this example.","section":"Example 5.2"},{"comment":"In the proof of Lemma 4.3, the notation for limit superior and limit inferior is visually ambiguous; please define the overline/underline notation explicitly.","section":"Lemma 4.3"}],"recommendation":"major_revision","confidential_remarks":"The paper's main theorem is internally consistent and the proof is largely sound, but the unproved infinite-dimensional assumption (28) is load-bearing for the advertised scope. The authors will need to either prove this identity or present the result as conditional in a way that is clearly reflected in the abstract and introduction. The heavy reliance on the authors' prior results is acceptable for this research program, but the novelty may be more incremental if the infinite-dimensional part is weakened."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nThe short version: this is a real advance in the theory of Douglas–Rachford splitting, and the proof is careful. Bauschke and Moursi prove weak convergence of the shadow sequence (P_U T^n x) to a minimizer of the shifted objective ι_U + g(·−v) for the inconsistent problem of minimizing a convex function subject to a linear subspace constraint. That result, Theorem 5.1, is new: earlier work covered only indicator functions or did not give iterate convergence. The paper also gives a parallel splitting application, and the examples usefully show which assumptions are necessary.\n\nWhat I like: the function-value analysis (Lemmas 4.2 and 4.3) is genuine, the identification Z = argmin(ι_U + g(·−v)) in Theorem 3.4 is new and central, and the paper does not hide its assumptions. Example 5.5 demonstrates that the constraint qualification (11) is needed, and Example 3.6 shows Z nonempty is not automatic. The self-citations are to a coherent program, and the quoted facts are standard or proved elsewhere; I do not see circularity.\n\nThe soft spot is assumption (28): v = P_{U−dom g}(0). Proposition 3.1 proves it in finite dimensions but the paper simply assumes it for general Hilbert space. This is load-bearing, since (29), v ∈ U⊥, is used in Lemma 4.1 and Proposition 3.9 to reach Theorem 5.1. The paper flags this clearly, so it is not a hidden flaw, but the reader should know the main theorem is conditional on an unproved identity in infinite dimensions. The stress-test note asks whether (11) alone would guarantee (28); the paper does not answer that, and Example 5.5 is not a counterexample because it violates (11). So the infinite-dimensional scope is narrower than the abstract alone suggests. This is a limitation, not a fatal one.\n\nWho is this for? Convex optimization researchers interested in operator splitting for infeasible or inconsistent problems. The result is significant enough to justify a careful referee. My own verdict is accept after minor revision; I would ask the authors to either prove (28) under weaker conditions or state explicitly that it remains open in infinite dimensions.\n\nThe paper deserves a serious referee.","headline":"A genuine advance in Douglas-Rachford theory for inconsistent convex problems, with a clearly flagged but load-bearing infinite-dimensional assumption that slightly narrows the stated scope.","tokens_in":18365,"tokens_out":2543,"would_cite":true,"duration_ms":24429,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["49M27","65K10","90C25","47H14","49M29"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper proves that the Douglas–Rachford algorithm converges weakly to a minimizer of a shifted convex program even when the original constrained problem has no solution.","keywords":["convex optimization","Douglas-Rachford algorithm","inconsistent constrained optimization","normal solution","minimal displacement vector","parallel splitting method","projection operator","proximal mapping"],"falsifier":"Find an infinite-dimensional Hilbert space, a closed linear subspace $U$, and a convex lower semicontinuous proper $g$ for which $v=P_{U-\\operatorname{dom}g}(0)$, $Z\\neq\\varnothing$, and some starting point $x$ produces a shadow sequence $P_UT^n x$ with a weak cluster point outside $\\arg\\min(\\iota_U+g(\\cdot-v))$; Theorem 5.1 declares this impossible. A concrete low-cost check is the paper's Example 5.3, where the formula yields $P_UT^n x=0$ for every $n$; a symbolic or numerical run producing any other limit would indicate a failure of the identity $P_UP_F=P_Z$.","tokens_in":17382,"feed_emoji":"🎯","tokens_out":19705,"duration_ms":165330,"temperature":0.7,"pith_summary":"The Douglas–Rachford algorithm is a standard splitting method for minimizing the sum of two convex functions; here the two functions are the indicator of a closed linear subspace $U$ and a general convex function $g$. The paper's aim is to understand what the algorithm still does when the problem $\\min_{x\\in X}(\\iota_U(x)+g(x))$ has no solution, because $\\operatorname{dom}g$ and $U$ may be disjoint. It proves that if the gap vector $v=P_{U-\\operatorname{dom}g}(0)$ is used to shift $g$, then the shadow sequence $P_UT^n x$ (the projection of the iterates onto $U$) converges weakly to a point of $\\arg\\min(\\iota_U+g(\\cdot-v))$, and the function values $g(P_g R_U T^n x)$ converge to the infimum of the shifted problem. A sympathetic reader should care because inconsistent constraints appear naturally in feasibility and parallel-splitting problems, and this result shows the algorithm still produces a meaningful normal solution rather than diverging arbitrarily. In particular, it yields a new parallel splitting theorem for minimizing a sum of convex functions under possibly inconsistent constraints.","feed_headline":"Douglas-Rachford converges even when no solution exists","feed_subtitle":"A gap vector and a shifted objective turn inconsistent linear constraints into a solvable normal problem.","key_machinery":"The machinery is the minimal displacement vector $v=P_{U-\\operatorname{dom}g}(0)$, together with the identity $Z=\\arg\\min(\\iota_U+g(\\cdot-v))$ for the normal-solution set $Z=\\{x:v\\in N_U(x)+\\partial g(x-v)\\}$. The vector $v$, which is shown to lie in $U^\\perp$, converts an inconsistent problem into a consistent shifted one. The proof is carried by the Douglas–Rachford operator $T=\\operatorname{Id}-P_U+P_gR_U$ with reflector $R_U=2P_U-\\operatorname{Id}$, the generalized fixed-point set $F=\\operatorname{Fix}T(\\cdot+v)$, and the projection relation $P_UP_F=P_Z$. Once these static identities are in place, a function-value analysis shows that the prox terms $P_gR_UT^n x$ have all weak cluster points in $\\arg\\min(\\iota_{U-v}+g)$ and that $g(P_gR_UT^n x)$ converges to the shifted infimum; Theorem 5.1 then lifts this to weak convergence of the shadow sequence.","core_discovery":"The central claim, stated in Theorem 5.1, is that for every starting point $x$, $P_UT^n x\\rightharpoonup P_Uy(x)\\in\\arg\\min(\\iota_U+g(\\cdot-v))$ and $g(P_gR_UT^n x)\\to\\min(\\iota_U+g(\\cdot-v))$, where $y(x)=\\lim_{n\\to\\infty}P_F(nv+T^n x)$ and $F=\\operatorname{Fix}T(\\cdot+v)$. In words: even though the original objective $\\iota_U+g$ may have infimum $+\\infty$, the algorithm converges to a minimizer of the minimally shifted objective $\\iota_U+g(\\cdot-v)$. The shift $v$ is the projection of $0$ onto $U-\\operatorname{dom}g$ and belongs to $U^\\perp$, so it measures the geometric gap between the constraint space and the function's domain. The proof works by identifying the normal-solution set $Z=\\{x:v\\in N_U(x)+\\partial g(x-v)\\}$ with $\\arg\\min(\\iota_U+g(\\cdot-v))$ and by using the projection identity $P_UP_F=P_Z$ to show that every weak cluster point of the shadow sequence is the same point, namely $P_Uy(x)$.","pith_inferences":["Editorial inference: the proved convergence $g(P_gR_UT^n x)\\to\\min(\\iota_U+g(\\cdot-v))$ suggests a practical stopping rule based on successive function values; the paper explicitly leaves termination criteria and numerical experiments for future research (Remark 5.7).","Editorial inference: because $v$ is characterized as the limit of $(P_U-\\operatorname{Id})P_{\\operatorname{dom}g}P_U$ (Fact 2.2), one could estimate $v$ on the fly during the iteration and then switch to the shifted problem; this two-phase procedure is not described in the paper.","Editorial inference: the proof uses linearity of $U$ in places such as $P_CP_U=P_C$ for $C\\subseteq U$, so extending the result to a general closed convex constraint set would require a new argument; a low-cost first check is to reproduce the closed-form predictions of Example 5.3 numerically.","Editorial inference: in the parallel-splitting setting, each shift $v_i$ can be read as a measure of how much the $i$-th constraint must be relaxed to make the system consistent; this interpretation is a direct reading of Corollary 6.7, though the paper does not spell it out."],"forward_implications":["If the assumptions hold, the algorithm finds a normal solution even when $U\\cap\\operatorname{dom}g=\\varnothing$: the shadow sequence converges weakly to a minimizer of $\\iota_U+g(\\cdot-v)$, and the function values $g(P_gR_UT^n x)$ converge to the infimum of $g$ over $U-v$.","The identity $Z=\\arg\\min(\\iota_U+g(\\cdot-v))$ gives the abstract normal solution an explicit interpretation as the solution of an ordinary shifted convex program, making the normal problem computationally meaningful.","For the sum of finitely many convex functions, the product-space formulation (Corollary 6.7) gives convergence of the parallel Douglas-Rachford updates to a minimizer of $\\sum_i g_i(\\cdot-v_i)$; functions with full domain are unshifted.","When $g=\\iota_W$ is an indicator, the result reduces to known affine-convex and two-set feasibility behavior: the shadow sequence converges to a point of $U\\cap(v+W)$.","The theorem covers cases not handled by earlier infinite-dimensional two-indicator results, since it does not require $Z\\subseteq F$; Example 5.3 has $Z\\cap F=\\varnothing$ yet the shadow sequence converges."],"supporting_citations":[{"why":"Introduces the normal problem and the set Z of normal solutions; the paper's target set and nonemptiness assumption come from here.","marker":"[8]"},{"why":"Provides the first function-value analysis for pathological convex programs and proves the finite-dimensional form of the gap vector formula; the present paper extends this to Hilbert-space iterates.","marker":"[25]"},{"why":"Gives the finite-dimensional description of v through the range of the Douglas-Rachford operator, used to justify assumption (28).","marker":"[9]"},{"why":"Supplies Fact 2.2, showing that the projection of 0 onto U-C lies in U^perp; this yields the crucial membership v in U^perp.","marker":"[2]"},{"why":"Proves shadow-sequence convergence for two indicator functions in infinite dimensions under Z subset of F; the paper notes Theorem 5.1 removes that inclusion.","marker":"[12]"},{"why":"Analyzes the affine-subspace case with strong convergence and supplies operator facts used in Fact 2.1.","marker":"[11]"},{"why":"Defines the Douglas-Rachford splitting algorithm whose behavior is the subject of the paper.","marker":"[16]"},{"why":"Establishes the classical weak convergence of the shadow sequence in the consistent case; the present theorem is the inconsistent-case extension.","marker":"[26]"},{"why":"Standard convex-analysis reference for the projector, proximal mapping, and subdifferential facts used throughout.","marker":"[5]"}],"fun_headline_variants":["Douglas-Rachford converges even when constraints conflict","Shifted objective turns infeasible problem into solvable one","Normal solution found despite empty domain intersection","Douglas-Rachford handles inconsistent linear constraints","Gap vector guides Douglas-Rachford to normal solution"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the shift vector $v$ equals the projection of $0$ onto $U-\\operatorname{dom}g$—automatic in finite dimensions but assumed in general Hilbert spaces—along with the assumption that the normal-solution set $Z$ is nonempty; without these, the identity $Z=\\arg\\min(\\iota_U+g(\\cdot-v))$ and the weak convergence of the shadow sequence are not established.","fun_headline_variants_meta":{"raw":{"variants":["Douglas-Rachford converges even when constraints conflict","Shifted objective turns infeasible problem into solvable one","Normal solution found despite empty domain intersection","Douglas-Rachford handles inconsistent linear constraints","Gap vector guides Douglas-Rachford to normal solution"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000203,"raw_usage":{"total_tokens":1383,"prompt_tokens":938,"completion_tokens":445,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":554,"completion_tokens_details":{"reasoning_tokens":372}},"tokens_in":554,"tokens_out":445,"duration_ms":5022,"temperature":1.0,"reasoning_tokens":372,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:15:09.293714+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Find an infinite-dimensional Hilbert space, a closed linear subspace $U$, and a convex lower semicontinuous proper $g$ for which $v=P_{U-\\operatorname{dom}g}(0)$, $Z\\neq\\varnothing$, and some starting point $x$ produces a shadow sequence $P_UT^n x$ with a weak cluster point outside $\\arg\\min(\\iota_U+g(\\cdot-v))$; Theorem 5.1 declares this impossible. A concrete low-cost check is the paper's Example 5.3, where the formula yields $P_UT^n x=0$ for every $n$; a symbolic or numerical run producing any other limit would indicate a failure of the identity $P_UP_F=P_Z$.","supporting_citations":[{"cited_title":"Bauschke, W.L","cited_arxiv_id":null,"evidence_quote":"Introduces the normal problem and the set Z of normal solutions; the paper's target set and nonemptiness assumption come from here."},{"cited_title":"Douglas--Rachford Splitting and ADMM for Pathological Convex Optimization","cited_arxiv_id":"1801.06618","evidence_quote":"Provides the first function-value analysis for pathological convex programs and proves the finite-dimensional form of the gap vector formula; the present paper extends this to Hilbert-space iterates."},{"cited_title":"Bauschke, W.L","cited_arxiv_id":null,"evidence_quote":"Gives the finite-dimensional description of v through the range of the Douglas-Rachford operator, used to justify assumption (28)."},{"cited_title":"Bauschke and J.M","cited_arxiv_id":null,"evidence_quote":"Supplies Fact 2.2, showing that the projection of 0 onto U-C lies in U^perp; this yields the crucial membership v in U^perp."},{"cited_title":"Bauschke and W.M","cited_arxiv_id":null,"evidence_quote":"Proves shadow-sequence convergence for two indicator functions in infinite dimensions under Z subset of F; the paper notes Theorem 5.1 removes that inclusion."},{"cited_title":"Bauschke and W.M","cited_arxiv_id":null,"evidence_quote":"Analyzes the affine-subspace case with strong convergence and supplies operator facts used in Fact 2.1."},{"cited_title":"Douglas and H.H","cited_arxiv_id":null,"evidence_quote":"Defines the Douglas-Rachford splitting algorithm whose behavior is the subject of the paper."},{"cited_title":"Svaiter, On weak convergence of the Douglas-Rachf ord method, SIAM Journal on Control and Optimization 49 (2011), 280–287","cited_arxiv_id":null,"evidence_quote":"Establishes the classical weak convergence of the shadow sequence in the consistent case; the present theorem is the inconsistent-case extension."},{"cited_title":"Bauschke and P .L","cited_arxiv_id":null,"evidence_quote":"Standard convex-analysis reference for the projector, proximal mapping, and subdifferential facts used throughout."}],"review_version":1}