# \#sddp

**URL:** https://discourse.julialang.org/tag/sddp/1460.md

[Latest](https://discourse.julialang.org/latest.md) · [Categories](https://discourse.julialang.org/categories.md) · [Tags](https://discourse.julialang.org/tags.md)

---

## [Question about SDDP.jl](https://discourse.julialang.org/t/question-about-sddp-jl/137741)

<div class="topic-metadata">

**Author:** [@david-hri](https://discourse.julialang.org/u/david-hri)\
**Replies:** 2\
**Last updated:** [June 26, 2026, 8:15am UTC](https://discourse.julialang.org/t/question-about-sddp-jl/137741 "2026-06-26T08:15:57Z")

</div>

Hello, I am new to SDDP.jl, so my question may be naive. I currently have a working SDDP model in which the generation profits are computed using a fixed hydraulic head effect (so the efficiency does not depend on the…

---

## [Markovian SDDP: is a finite and concrete scenario tree necessary?](https://discourse.julialang.org/t/markovian-sddp-is-a-finite-and-concrete-scenario-tree-necessary/134317)

<div class="topic-metadata">

**Author:** [@WalterMadelim](https://discourse.julialang.org/u/WalterMadelim)\
**Replies:** 12\
**Last updated:** [December 13, 2025, 3:27am UTC](https://discourse.julialang.org/t/markovian-sddp-is-a-finite-and-concrete-scenario-tree-necessary/134317 "2025-12-13T03:27:14Z")

</div>

(I kinda forget these knowledge) I asked chatGPT this question is it a prerequisite to restrict the number of uncertainty realizations in each stage to be finite, if I adopt a Markov-SDDP algorithm to tackle a multist…

---

## [Sensitivity analysis in simulation phase of SDDP.jl](https://discourse.julialang.org/t/sensitivity-analysis-in-simulation-phase-of-sddp-jl/132085)

<div class="topic-metadata">

**Author:** [@Sorooshsa](https://discourse.julialang.org/u/Sorooshsa)\
**Replies:** 1\
**Last updated:** [September 4, 2025, 1:06am UTC](https://discourse.julialang.org/t/sensitivity-analysis-in-simulation-phase-of-sddp-jl/132085 "2025-09-04T01:06:46Z")

</div>

Hello, I trained a hydropower model using SDDP.jl with a given set of parameters. To check the robustness of the trained policy, I would now like to run simulations where one of the parameters is changed — for example, …

---

## [Rigorous establishment of the feasibility of the trained policies in SDDP.jl using constraint violation penalties in the objective function](https://discourse.julialang.org/t/rigorous-establishment-of-the-feasibility-of-the-trained-policies-in-sddp-jl-using-constraint-violation-penalties-in-the-objective-function/130961)

<div class="topic-metadata">

**Author:** [@Engr\_Moiz\_Ahmad](https://discourse.julialang.org/u/Engr_Moiz_Ahmad)\
**Replies:** 10\
**Last updated:** [July 23, 2025, 8:32am UTC](https://discourse.julialang.org/t/rigorous-establishment-of-the-feasibility-of-the-trained-policies-in-sddp-jl-using-constraint-violation-penalties-in-the-objective-function/130961 "2025-07-23T08:32:16Z")

</div>

I am working with a multi-stage mixed-integer stochastic global supply chain optimization problem. In this regard, under realistic parameter settings, some constraints are making the instances violate the relatively com…

---

## [Not achieving convergence after prolonged training](https://discourse.julialang.org/t/not-achieving-convergence-after-prolonged-training/130093)

<div class="topic-metadata">

**Author:** [@Engr\_Moiz\_Ahmad](https://discourse.julialang.org/u/Engr_Moiz_Ahmad)\
**Replies:** 3\
**Last updated:** [June 22, 2025, 8:03pm UTC](https://discourse.julialang.org/t/not-achieving-convergence-after-prolonged-training/130093 "2025-06-22T20:03:21Z")

</div>

I am dealing with an infinite-horizon multi-stage mixed-integer linear stochastic global supply chain optimization problem and using SDDP.jl for its solution. In this regard, after a significant amount of training, the …

---

## [For a training iteration or the evaluation simulation (after training), is burn-in period implemented in SDDP.jl?](https://discourse.julialang.org/t/for-a-training-iteration-or-the-evaluation-simulation-after-training-is-burn-in-period-implemented-in-sddp-jl/129652)

<div class="topic-metadata">

**Author:** [@Engr\_Moiz\_Ahmad](https://discourse.julialang.org/u/Engr_Moiz_Ahmad)\
**Replies:** 4\
**Last updated:** [June 6, 2025, 9:35pm UTC](https://discourse.julialang.org/t/for-a-training-iteration-or-the-evaluation-simulation-after-training-is-burn-in-period-implemented-in-sddp-jl/129652 "2025-06-06T21:35:11Z")

</div>

I am dealing with a multi-stage mixed-integer stochastic global supply chain optimization problem. In this regard, I want to determine whether the burn-in period (for the random sampling process) is already incorporated…

---

## [What should be the acceptable tolerance value between the lower bound and the estimated upper bound for the results of SDDP.jl to be reliable?](https://discourse.julialang.org/t/what-should-be-the-acceptable-tolerance-value-between-the-lower-bound-and-the-estimated-upper-bound-for-the-results-of-sddp-jl-to-be-reliable/129450)

<div class="topic-metadata">

**Author:** [@Engr\_Moiz\_Ahmad](https://discourse.julialang.org/u/Engr_Moiz_Ahmad)\
**Replies:** 4\
**Last updated:** [May 30, 2025, 7:40am UTC](https://discourse.julialang.org/t/what-should-be-the-acceptable-tolerance-value-between-the-lower-bound-and-the-estimated-upper-bound-for-the-results-of-sddp-jl-to-be-reliable/129450 "2025-05-30T07:40:04Z")

</div>

I am dealing with a multi-stage mixed-integer stochastic global supply chain optimization problem. In this regard, I want to determine a tolerance value for the gap between the lower bound and the estimated upper bound …

---

## [How to solve infinite horizon problem under Makovian policy graph in SDDP.jl?](https://discourse.julialang.org/t/how-to-solve-infinite-horizon-problem-under-makovian-policy-graph-in-sddp-jl/129420)

<div class="topic-metadata">

**Author:** [@Engr\_Moiz\_Ahmad](https://discourse.julialang.org/u/Engr_Moiz_Ahmad)\
**Replies:** 16\
**Last updated:** [May 29, 2025, 7:28pm UTC](https://discourse.julialang.org/t/how-to-solve-infinite-horizon-problem-under-makovian-policy-graph-in-sddp-jl/129420 "2025-05-29T19:28:08Z")

</div>

I am dealing with a multi-stage mixed-integer stochastic global supply chain optimization problem and want to solve infinite-horizon problem instances under a Markovian policy graph. I am using the Markovian policy graph…

---

## [Do degeneracies and numerical instabilities more likely during the execution of solvers in case of formulations with greater number of constraints and decision variables?](https://discourse.julialang.org/t/do-degeneracies-and-numerical-instabilities-more-likely-during-the-execution-of-solvers-in-case-of-formulations-with-greater-number-of-constraints-and-decision-variables/129108)

<div class="topic-metadata">

**Author:** [@Engr\_Moiz\_Ahmad](https://discourse.julialang.org/u/Engr_Moiz_Ahmad)\
**Replies:** 2\
**Last updated:** [May 18, 2025, 2:23am UTC](https://discourse.julialang.org/t/do-degeneracies-and-numerical-instabilities-more-likely-during-the-execution-of-solvers-in-case-of-formulations-with-greater-number-of-constraints-and-decision-variables/129108 "2025-05-18T02:23:58Z")

</div>

I am dealing with a multi-stage mixed-integer stochastic global supply chain optimization problem and want to solve it using the SDDiP algorithm in the SDDP.jl package in Julia. In this regard, I am conducting prelimina…

---

## [Unable to solve the problem using Lagrangian duality handler](https://discourse.julialang.org/t/unable-to-solve-the-problem-using-lagrangian-duality-handler/129079)

<div class="topic-metadata">

**Author:** [@Engr\_Moiz\_Ahmad](https://discourse.julialang.org/u/Engr_Moiz_Ahmad)\
**Replies:** 2\
**Last updated:** [May 17, 2025, 9:40pm UTC](https://discourse.julialang.org/t/unable-to-solve-the-problem-using-lagrangian-duality-handler/129079 "2025-05-17T21:40:17Z")

</div>

I am dealing with a multi-stage mixed-integer stochastic global supply chain optimization problem and want to solve it using the SDDiP algorithm in the SDDP.jl package in Julia. In this regard, I tried solving the mixed…

---

## [Why convergence behavior of SDDiP algorithm is different with different mixed-integer programming solvers?](https://discourse.julialang.org/t/why-convergence-behavior-of-sddip-algorithm-is-different-with-different-mixed-integer-programming-solvers/128913)

<div class="topic-metadata">

**Author:** [@Engr\_Moiz\_Ahmad](https://discourse.julialang.org/u/Engr_Moiz_Ahmad)\
**Replies:** 11\
**Last updated:** [May 15, 2025, 1:45am UTC](https://discourse.julialang.org/t/why-convergence-behavior-of-sddip-algorithm-is-different-with-different-mixed-integer-programming-solvers/128913 "2025-05-15T01:45:49Z")

</div>

I am dealing with a multi-stage mixed-integer stochastic global supply chain optimization problem and want to solve it using the SDDiP algorithm using the SDDP.jl package in Julia. In this regard, I tried various mixed-…

---

## [How to save a trained policy in SDDP.jl?](https://discourse.julialang.org/t/how-to-save-a-trained-policy-in-sddp-jl/128894)

<div class="topic-metadata">

**Author:** [@Engr\_Moiz\_Ahmad](https://discourse.julialang.org/u/Engr_Moiz_Ahmad)\
**Replies:** 2\
**Last updated:** [May 11, 2025, 5:48pm UTC](https://discourse.julialang.org/t/how-to-save-a-trained-policy-in-sddp-jl/128894 "2025-05-11T17:48:37Z")

</div>

I am dealing with a multi-stage mixed-integer stochastic global supply chain optimization problem and want to solve it using the SDDiP algorithm using the SDDP.jl package in Julia. In this regard, I want to download the…

---

## [Best way to train and evaluate SDDP with constraints including product of control and states variables?](https://discourse.julialang.org/t/best-way-to-train-and-evaluate-sddp-with-constraints-including-product-of-control-and-states-variables/128890)

<div class="topic-metadata">

**Author:** [@petr-a](https://discourse.julialang.org/u/petr-a)\
**Replies:** 1\
**Last updated:** [May 10, 2025, 11:27pm UTC](https://discourse.julialang.org/t/best-way-to-train-and-evaluate-sddp-with-constraints-including-product-of-control-and-states-variables/128890 "2025-05-10T23:27:30Z")

</div>

Hello! I’m solving a problem which is similar to the Milk producer example with the SDDP algorithm. This question is also related to my first question about SDDP and out-of-sample solutions. Here is my asset managemen…

---

## [Access node in custom\_recorders during historical simulation (SDDP.jl)](https://discourse.julialang.org/t/access-node-in-custom-recorders-during-historical-simulation-sddp-jl/128507)

<div class="topic-metadata">

**Author:** [@Sorooshsa](https://discourse.julialang.org/u/Sorooshsa)\
**Replies:** 1\
**Last updated:** [April 29, 2025, 12:40am UTC](https://discourse.julialang.org/t/access-node-in-custom-recorders-during-historical-simulation-sddp-jl/128507 "2025-04-29T00:40:14Z")

</div>

Hey, I am calculating the dual values of some constraints during historical simulation, and the constraint index depends on the node in the body of the problem. Therefore, in the simulation part, I need to refer to the n…

---

## [How does SDDP provide decision for an out-of-sample state?](https://discourse.julialang.org/t/how-does-sddp-provide-decision-for-an-out-of-sample-state/128433)

<div class="topic-metadata">

**Author:** [@petr-a](https://discourse.julialang.org/u/petr-a)\
**Replies:** 2\
**Last updated:** [April 28, 2025, 8:47am UTC](https://discourse.julialang.org/t/how-does-sddp-provide-decision-for-an-out-of-sample-state/128433 "2025-04-28T08:47:04Z")

</div>

Hello! I’m new to SDDP and Julia and I wonder how the policy evaluation works (math behind it) for states which are out of the scenario lattice which is used to train the algorithm? For context: I’m solving the problem…

---

## [Convergence of SDDP with nonlinear objective and mixed-integer variables](https://discourse.julialang.org/t/convergence-of-sddp-with-nonlinear-objective-and-mixed-integer-variables/127947)

<div class="topic-metadata">

**Author:** [@Engr\_Moiz\_Ahmad](https://discourse.julialang.org/u/Engr_Moiz_Ahmad)\
**Replies:** 9\
**Last updated:** [April 11, 2025, 7:55am UTC](https://discourse.julialang.org/t/convergence-of-sddp-with-nonlinear-objective-and-mixed-integer-variables/127947 "2025-04-11T07:55:30Z")

</div>

@odow what reference should I cite if I want to claim that due to the simultaneous occurrence of exponential terms in the objective and mixed integer decision space, SDDP can not be used?

---

## [How to train a biobjective problem with SDDP.jl?](https://discourse.julialang.org/t/how-to-train-a-biobjective-problem-with-sddp-jl/123590)

<div class="topic-metadata">

**Author:** [@z\_yt\_Yt](https://discourse.julialang.org/u/z_yt_Yt)\
**Replies:** 10\
**Last updated:** [January 14, 2025, 11:49pm UTC](https://discourse.julialang.org/t/how-to-train-a-biobjective-problem-with-sddp-jl/123590 "2025-01-14T23:49:08Z")

</div>

Hello, I am trying to run this SDDP.jl/papers/biobjective\_sddp/BiObjectiveSDDP.jl at master · odow/SDDP.jl · GitHub and also other files in this folder SDDP.jl/papers/biobjective\_sddp at master · odow/SDDP.jl · GitHub. …

---

## [Considering water head in SDDP.jl](https://discourse.julialang.org/t/considering-water-head-in-sddp-jl/119779)

<div class="topic-metadata">

**Author:** [@Sorooshsa](https://discourse.julialang.org/u/Sorooshsa)\
**Replies:** 1\
**Last updated:** [September 24, 2024, 3:34am UTC](https://discourse.julialang.org/t/considering-water-head-in-sddp-jl/119779 "2024-09-24T03:34:14Z")

</div>

Hi everyone, I’m using SDDP.jl to model hydropower production and want to include the effect of water head in the production constraint, which is typically formulated as P = release×water head×efficiency parameter. He…

---

## [Historical data - SDDP.jl](https://discourse.julialang.org/t/historical-data-sddp-jl/117207)

<div class="topic-metadata">

**Author:** [@Sorooshsa](https://discourse.julialang.org/u/Sorooshsa)\
**Replies:** 6\
**Last updated:** [August 6, 2024, 11:42pm UTC](https://discourse.julialang.org/t/historical-data-sddp-jl/117207 "2024-08-06T23:42:24Z")

</div>

Hi everyone, I trained a big model using SDDP.jl. Then, I used historical data to simulate the policy multiple times. However, when I examine the results of different simulations with the same historical data, I notice …

---

## [Different objective state functions for different stages SDDP.jl](https://discourse.julialang.org/t/different-objective-state-functions-for-different-stages-sddp-jl/116812)

<div class="topic-metadata">

**Author:** [@tumpelopumpelo](https://discourse.julialang.org/u/tumpelopumpelo)\
**Replies:** 3\
**Last updated:** [July 22, 2024, 7:31am UTC](https://discourse.julialang.org/t/different-objective-state-functions-for-different-stages-sddp-jl/116812 "2024-07-22T07:31:17Z")

</div>

Hi, I am quite new to optimization with SDDP so may be a silly question. I am working on optimising a bidding strategy for a wind power producer in consecutive markets (day ahead, Intraday, ancillary) using SDDP, with u…

---

## [Cyclic policy graph with fix number of cycles - SDDP.jl](https://discourse.julialang.org/t/cyclic-policy-graph-with-fix-number-of-cycles-sddp-jl/116546)

<div class="topic-metadata">

**Author:** [@Sorooshsa](https://discourse.julialang.org/u/Sorooshsa)\
**Replies:** 1\
**Last updated:** [July 13, 2024, 12:50am UTC](https://discourse.julialang.org/t/cyclic-policy-graph-with-fix-number-of-cycles-sddp-jl/116546 "2024-07-13T00:50:05Z")

</div>

Hi everyone, Can we design a cyclic linear policy graph in SDDP.jl with a fixed number of cycles? Let’s assume there are 4 nodes, where node 4 transitions to node 1 with a probability of 1. However, this cycle should re…

---

## [Why can't I set the \`initial\_value\` to a matrix value when using \`SDDP.jl\`?](https://discourse.julialang.org/t/why-cant-i-set-the-initial-value-to-a-matrix-value-when-using-sddp-jl/116086)

<div class="topic-metadata">

**Author:** [@WuSiren](https://discourse.julialang.org/u/WuSiren)\
**Replies:** 2\
**Last updated:** [June 23, 2024, 1:44am UTC](https://discourse.julialang.org/t/why-cant-i-set-the-initial-value-to-a-matrix-value-when-using-sddp-jl/116086 "2024-06-23T01:44:25Z")

</div>

In the hydro-thermal scheduling example, if I define a matrix type state variable, I will not be able to set a matrix initial value to it: using SDDP graph = SDDP.LinearGraph(3) function subproblem\_builder(subproblem:…

---

## [Co-optimizing renewables - SDDP.jl - Policy Graph](https://discourse.julialang.org/t/co-optimizing-renewables-sddp-jl-policy-graph/115289)

<div class="topic-metadata">

**Author:** [@Sorooshsa](https://discourse.julialang.org/u/Sorooshsa)\
**Replies:** 6\
**Last updated:** [June 17, 2024, 9:02pm UTC](https://discourse.julialang.org/t/co-optimizing-renewables-sddp-jl-policy-graph/115289 "2024-06-17T21:02:12Z")

</div>

Hi, I am working on co-optimizing renewables (wind and hydro) expansion and generation while accounting for uncertainty in the generation part. In these problems, one of the common issues is non-linearity in the constra…

---

## [End-of-horizon effect, Infinite horizon policy graph, SDDP.jl](https://discourse.julialang.org/t/end-of-horizon-effect-infinite-horizon-policy-graph-sddp-jl/115573)

<div class="topic-metadata">

**Author:** [@Sorooshsa](https://discourse.julialang.org/u/Sorooshsa)\
**Replies:** 11\
**Last updated:** [June 14, 2024, 12:52am UTC](https://discourse.julialang.org/t/end-of-horizon-effect-infinite-horizon-policy-graph-sddp-jl/115573 "2024-06-14T00:52:04Z")

</div>

Hi everyone, Let’s assume a simple co-optimized expansion and generation problem with 13 nodes: 1 for expansion and 12 generations corresponding to 12 months. There is uncertainty during generation, e.g., monthly water …

---

## [Decision-dependent randomness in SDDP.jl](https://discourse.julialang.org/t/decision-dependent-randomness-in-sddp-jl/114625)

<div class="topic-metadata">

**Author:** [@mothmani](https://discourse.julialang.org/u/mothmani)\
**Replies:** 6\
**Last updated:** [May 26, 2024, 1:45pm UTC](https://discourse.julialang.org/t/decision-dependent-randomness-in-sddp-jl/114625 "2024-05-26T13:45:16Z")

</div>

Hi everyone, I’m quite new to Julia and SDDP and may ask a silly question. I’m trying to optimize the preventive replacement of some devices over several stages. I assume that know the lifetime distribution of the devi…

---

## [Custom recorder on stages with different constraints](https://discourse.julialang.org/t/custom-recorder-on-stages-with-different-constraints/112718)

<div class="topic-metadata">

**Author:** [@FelixSchmidt](https://discourse.julialang.org/u/FelixSchmidt)\
**Replies:** 2\
**Last updated:** [April 9, 2024, 9:19pm UTC](https://discourse.julialang.org/t/custom-recorder-on-stages-with-different-constraints/112718 "2024-04-09T21:19:59Z")

</div>

Hi Oscar, I’ve been having issues with creating a custom recorder in SDDP.simulate. I have a capacity expansion problem where stage 1 covers the investment decision and stages 2,…T are monthly dispatch problems at hourl…

---

## [Marginal water value while using SDDP.jl](https://discourse.julialang.org/t/marginal-water-value-while-using-sddp-jl/112169)

<div class="topic-metadata">

**Author:** [@Sorooshsa](https://discourse.julialang.org/u/Sorooshsa)\
**Replies:** 7\
**Last updated:** [April 2, 2024, 12:24am UTC](https://discourse.julialang.org/t/marginal-water-value-while-using-sddp-jl/112169 "2024-04-02T00:24:54Z")

</div>

I have a multi-stage stochastic hydro-electric scheduling problem that SDDP.jl models and solves. The model includes a set of constraints related to water level control. So now I’d like to determine the reservoirs’ mar…

---

## [Accessing children nodes in the SDDP package](https://discourse.julialang.org/t/accessing-children-nodes-in-the-sddp-package/112140)

<div class="topic-metadata">

**Author:** [@heiwie](https://discourse.julialang.org/u/heiwie)\
**Replies:** 8\
**Last updated:** [March 28, 2024, 5:22pm UTC](https://discourse.julialang.org/t/accessing-children-nodes-in-the-sddp-package/112140 "2024-03-28T17:22:37Z")

</div>

I have an optimization problem where one of the constraints is a convex non-linear constraint on the following form: Here w, eta and x are decision variables. All variables with an m index are variables corresponding …

---

## [Indexing problem SDDP package](https://discourse.julialang.org/t/indexing-problem-sddp-package/110291)

<div class="topic-metadata">

**Author:** [@heiwie](https://discourse.julialang.org/u/heiwie)\
**Replies:** 4\
**Last updated:** [February 17, 2024, 10:33pm UTC](https://discourse.julialang.org/t/indexing-problem-sddp-package/110291 "2024-02-17T22:33:26Z")

</div>

I am trying to define a hydropower production problem in Julia using the SDDP package. When running the code I get this error: “ERROR: KeyError: key Symbol(“w2\[3\]”) not found”. This error seems like it is being caused b…

---

## [Historical simulation in SDDP.jl](https://discourse.julialang.org/t/historical-simulation-in-sddp-jl/109473)

<div class="topic-metadata">

**Author:** [@Sorooshsa](https://discourse.julialang.org/u/Sorooshsa)\
**Replies:** 3\
**Last updated:** [January 30, 2024, 10:01pm UTC](https://discourse.julialang.org/t/historical-simulation-in-sddp-jl/109473 "2024-01-30T22:01:33Z")

</div>

Hi, I did simulation in SDDP.jl using historical data. It works; however, I could not retrieve the optimal solutions anymore by simply putting \[:variable\] as an argument in SDDP.simulate. What should we do in historica…

[Next page](https://discourse.julialang.org/tag/sddp/1460.md?match_all_tags=true&page=1&tags%5B%5D=sddp)