Skip to content

Fixes for class 1 #14

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
andrewrosemberg merged 9 commits into from
Aug 14, 2025
Merged

Fixes for class 1 #14

Show file tree
Hide file tree
Changes from 4 commits
Commits
Show all changes
9 commits
Select commit Hold shift + click to select a range
4850c63
afine -> affine
klamike Aug 11, 2025
924caa4
pin 1.10.5
klamike Aug 11, 2025
c88010d
fix links
klamike Aug 11, 2025
ca2df6c
comment l2o org
klamike Aug 11, 2025
645816a
ylims
klamike Aug 12, 2025
f7109c6
hide in gui
klamike Aug 12, 2025
e5e1b32
gi manifests
klamike Aug 12, 2025
88f1679
unpin ci
klamike Aug 12, 2025
6039569
update
andrewrosemberg Aug 14, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion .github/workflows/documentation.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -17,7 +17,7 @@ jobs:
- uses: actions/checkout@v4
- uses: julia-actions/setup-julia@v2
with:
version: '1'
version: '1.10.5'
- uses: julia-actions/cache@v2
- name: Install dependencies
run: julia --project=docs/ -e 'using Pkg; Pkg.instantiate()'
Expand Down
286 changes: 143 additions & 143 deletions class01/class01_intro.jl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -57,13 +57,13 @@ begin
MarkdownLiteral.@markdown(
"""

[^cmu]: Zachary Manchester et al. Optimal Control and Reinforcement Learning at Carnegie Mellon University - [CMU 16-745]("https://optimalcontrol.ri.cmu.edu/")
[^cmu]: Zachary Manchester et al. [Optimal Control and Reinforcement Learning at Carnegie Mellon University - CMU 16-745](https://optimalcontrol.ri.cmu.edu/)

[^OptProx]: Van Hentenryck, P., 2024. Fusing Artificial Intelligence and Optimization with Trustworthy Optimization Proxies. Collections, 57(02).
[^OptProx]: Van Hentenryck, P., 2024. [Fusing Artificial Intelligence and Optimization with Trustworthy Optimization Proxies](https://www.siam.org/publications/siam-news/articles/fusing-artificial-intelligence-and-optimization-with-trustworthy-optimization-proxies/). Collections, 57(02).

[^ArmManip]: Guechi, E.H., Bouzoualegh, S., Zennir, Y. and Blažič, S., 2018. MPC control and LQ optimal control of a two-link robot arm: A comparative study. Machines, 6(3), p.37.
[^ArmManip]: Guechi, E.H., Bouzoualegh, S., Zennir, Y. and Blažič, S., 2018. [MPC control and LQ optimal control of a two-link robot arm: A comparative study](https://www.mdpi.com/2075-1702/6/3/37). Machines, 6(3), p.37.

[^ZachMIT]: Zachary Manchester talk at MIT - [MIT Robotics - Zac Manchester - Composable Optimization for Robotic Motion Planning and Control]("https://www.youtube.com/watch?v=eSleutHuc0w&ab_channel=MITRobotics").
[^ZachMIT]: Zachary Manchester talk at MIT - [MIT Robotics - Zac Manchester - Composable Optimization for Robotic Motion Planning and Control](https://www.youtube.com/watch?v=eSleutHuc0w&ab_channel=MITRobotics).

"""
)
Expand Down Expand Up @@ -210,143 +210,143 @@ md"[^OptProx]"

"""

# ╔═╡ 45275d44-e268-43cb-8156-feecd916a6da
@htl """
<div style="
border:1px solid #ccc;
border-radius:6px;
padding:1rem;
font-size:0.9rem;
max-width:760px;
line-height:1.45;
">

<!-- ─────────────────────── header ─────────────────────── -->
<h2 style="margin-top:0">LearningToOptimize&nbsp;Organization</h2>

<p>
<strong>LearningToOptimize&nbsp;(L2O)</strong> is a collection of open-source tools
focused on the emerging paradigm of <em>amortized optimization</em>—using machine-learning
methods to accelerate traditional constrained-optimization solvers.
<em>L2O is a work-in-progress; existing functionality is considered experimental and may
change.</em>
</p>

<!-- ─────────────────── repositories table ──────────────── -->
<h3>Open-Source&nbsp;Repositories</h3>

<table style="border-collapse:collapse;width:100%">
<tbody>
<tr>
<td style="padding:4px 6px;vertical-align:top;">
<a href="https://github.com/LearningToOptimize/LearningToOptimize.jl"
target="_blank">LearningToOptimize.jl</a>
</td>
<td style="padding:4px 6px;">
Flagship Julia package that wraps data generation, training loops and evaluation
utilities for fitting surrogate models to parametric optimization problems.
</td>
</tr>

<tr>
<td style="padding:4px 6px;vertical-align:top;">
<a href="https://github.com/andrewrosemberg/DecisionRules.jl"
target="_blank">DecisionRules.jl</a>
</td>
<td style="padding:4px 6px;">
Build decision rules for multistage stochastic programs, as proposed in
<a href="https://arxiv.org/pdf/2405.14973" target="_blank"><em>Efficiently
Training Deep-Learning Parametric Policies using Lagrangian Duality</em></a>.
</td>
</tr>

<tr>
<td style="padding:4px 6px;vertical-align:top;">
<a href="https://github.com/LearningToOptimize/L2OALM.jl"
target="_blank">L2OALM.jl</a>
</td>
<td style="padding:4px 6px;">
Implementation of the primal-dual learning method <strong>ALM</strong>,
introduced in
<a href="https://ojs.aaai.org/index.php/AAAI/article/view/25520" target="_blank">
<em>Self-Supervised Primal-Dual Learning for Constrained Optimization</em></a>.
</td>
</tr>

<tr>
<td style="padding:4px 6px;vertical-align:top;">
<a href="https://github.com/LearningToOptimize/L2ODLL.jl"
target="_blank">L2ODLL.jl</a>
</td>
<td style="padding:4px 6px;">
Implementation of the dual learning method <strong>DLL</strong>,
proposed in
<a href="https://neurips.cc/virtual/2024/poster/94146" target="_blank">
<em>Dual Lagrangian Learning for Conic Optimization</em></a>.
</td>
</tr>

<tr>
<td style="padding:4px 6px;vertical-align:top;">
<a href="https://github.com/LearningToOptimize/L2ODC3.jl"
target="_blank">L2ODC3.jl</a>
</td>
<td style="padding:4px 6px;">
Implementation of the primal learning method <strong>DC3</strong>, as described in
<a href="https://openreview.net/forum?id=V1ZHVxJ6dSS" target="_blank">
<em>DC3: A Learning Method for Optimization with Hard Constraints</em></a>.
</td>
</tr>

<tr>
<td style="padding:4px 6px;vertical-align:top;">
<a href="https://github.com/LearningToOptimize/BatchNLPKernels.jl"
target="_blank">BatchNLPKernels.jl</a>
</td>
<td style="padding:4px 6px;">
GPU kernels that evaluate objectives, Jacobians and Hessians for
<strong>batches</strong> of
<a href="https://github.com/exanauts/ExaModels.jl" target="_blank">ExaModels</a>,
useful when defining loss functions for large-batch ML predictions.
</td>
</tr>

<tr>
<td style="padding:4px 6px;vertical-align:top;">
<a href="https://github.com/LearningToOptimize/BatchConeKernels.jl"
target="_blank">BatchConeKernels.jl</a>
</td>
<td style="padding:4px 6px;">
GPU kernels for batched cone operations (projections, distances, etc.),
enabling advanced architectures such as repair layers.
</td>
</tr>

<tr>
<td style="padding:4px 6px;vertical-align:top;">
<a href="https://github.com/LearningToOptimize/LearningToControlClass"
target="_blank">LearningToControlClass</a>
</td>
<td style="padding:4px 6px;">
Course repository for <em>Special Topics on Optimal Control &amp; Learning</em>
(Fall 2025, Georgia Tech).
</td>
</tr>
</tbody>
</table>

<!-- ─────────────── datasets and weights ──────────────── -->
<h3 style="margin-top:1.25rem;">Open Datasets and Weights</h3>

<p>
The
<a href="https://huggingface.co/LearningToOptimize" target="_blank">
LearningToOptimize&nbsp;🤗 Hugging Face organization</a>
hosts datasets and pre-trained weights that can be used with L2O packages.
</p>

</div>
"""
# # ╔═╡ 45275d44-e268-43cb-8156-feecd916a6da
# @htl """
# <div style="
# border:1px solid #ccc;
# border-radius:6px;
# padding:1rem;
# font-size:0.9rem;
# max-width:760px;
# line-height:1.45;
# ">

# <!-- ─────────────────────── header ─────────────────────── -->
# <h2 style="margin-top:0">LearningToOptimize&nbsp;Organization</h2>

# <p>
# <strong>LearningToOptimize&nbsp;(L2O)</strong> is a collection of open-source tools
# focused on the emerging paradigm of <em>amortized optimization</em>—using machine-learning
# methods to accelerate traditional constrained-optimization solvers.
# <em>L2O is a work-in-progress; existing functionality is considered experimental and may
# change.</em>
# </p>

# <!-- ─────────────────── repositories table ──────────────── -->
# <h3>Open-Source&nbsp;Repositories</h3>

# <table style="border-collapse:collapse;width:100%">
# <tbody>
# <tr>
# <td style="padding:4px 6px;vertical-align:top;">
# <a href="https://github.com/LearningToOptimize/LearningToOptimize.jl"
# target="_blank">LearningToOptimize.jl</a>
# </td>
# <td style="padding:4px 6px;">
# Flagship Julia package that wraps data generation, training loops and evaluation
# utilities for fitting surrogate models to parametric optimization problems.
# </td>
# </tr>

# <tr>
# <td style="padding:4px 6px;vertical-align:top;">
# <a href="https://github.com/andrewrosemberg/DecisionRules.jl"
# target="_blank">DecisionRules.jl</a>
# </td>
# <td style="padding:4px 6px;">
# Build decision rules for multistage stochastic programs, as proposed in
# <a href="https://arxiv.org/pdf/2405.14973" target="_blank"><em>Efficiently
# Training Deep-Learning Parametric Policies using Lagrangian Duality</em></a>.
# </td>
# </tr>

# <tr>
# <td style="padding:4px 6px;vertical-align:top;">
# <a href="https://github.com/LearningToOptimize/L2OALM.jl"
# target="_blank">L2OALM.jl</a>
# </td>
# <td style="padding:4px 6px;">
# Implementation of the primal-dual learning method <strong>ALM</strong>,
# introduced in
# <a href="https://ojs.aaai.org/index.php/AAAI/article/view/25520" target="_blank">
# <em>Self-Supervised Primal-Dual Learning for Constrained Optimization</em></a>.
# </td>
# </tr>

# <tr>
# <td style="padding:4px 6px;vertical-align:top;">
# <a href="https://github.com/LearningToOptimize/L2ODLL.jl"
# target="_blank">L2ODLL.jl</a>
# </td>
# <td style="padding:4px 6px;">
# Implementation of the dual learning method <strong>DLL</strong>,
# proposed in
# <a href="https://neurips.cc/virtual/2024/poster/94146" target="_blank">
# <em>Dual Lagrangian Learning for Conic Optimization</em></a>.
# </td>
# </tr>

# <tr>
# <td style="padding:4px 6px;vertical-align:top;">
# <a href="https://github.com/LearningToOptimize/L2ODC3.jl"
# target="_blank">L2ODC3.jl</a>
# </td>
# <td style="padding:4px 6px;">
# Implementation of the primal learning method <strong>DC3</strong>, as described in
# <a href="https://openreview.net/forum?id=V1ZHVxJ6dSS" target="_blank">
# <em>DC3: A Learning Method for Optimization with Hard Constraints</em></a>.
# </td>
# </tr>

# <tr>
# <td style="padding:4px 6px;vertical-align:top;">
# <a href="https://github.com/LearningToOptimize/BatchNLPKernels.jl"
# target="_blank">BatchNLPKernels.jl</a>
# </td>
# <td style="padding:4px 6px;">
# GPU kernels that evaluate objectives, Jacobians and Hessians for
# <strong>batches</strong> of
# <a href="https://github.com/exanauts/ExaModels.jl" target="_blank">ExaModels</a>,
# useful when defining loss functions for large-batch ML predictions.
# </td>
# </tr>

# <tr>
# <td style="padding:4px 6px;vertical-align:top;">
# <a href="https://github.com/LearningToOptimize/BatchConeKernels.jl"
# target="_blank">BatchConeKernels.jl</a>
# </td>
# <td style="padding:4px 6px;">
# GPU kernels for batched cone operations (projections, distances, etc.),
# enabling advanced architectures such as repair layers.
# </td>
# </tr>

# <tr>
# <td style="padding:4px 6px;vertical-align:top;">
# <a href="https://github.com/LearningToOptimize/LearningToControlClass"
# target="_blank">LearningToControlClass</a>
# </td>
# <td style="padding:4px 6px;">
# Course repository for <em>Special Topics on Optimal Control &amp; Learning</em>
# (Fall 2025, Georgia Tech).
# </td>
# </tr>
# </tbody>
# </table>

# <!-- ─────────────── datasets and weights ──────────────── -->
# <h3 style="margin-top:1.25rem;">Open Datasets and Weights</h3>

# <p>
# The
# <a href="https://huggingface.co/LearningToOptimize" target="_blank">
# LearningToOptimize&nbsp;🤗 Hugging Face organization</a>
# hosts datasets and pre-trained weights that can be used with L2O packages.
# </p>

# </div>
# """

# ╔═╡ c08f511e-b91d-4d17-a286-96469c31568a
md"## Example: Robotic Arm Manipulation"
Expand Down Expand Up @@ -585,7 +585,7 @@ x \in S^{1} \times \mathbb{R} & \text{Cylinder}
""")

# ╔═╡ 4d598933-05a9-44fa-b5a7-f7e1c7afb094
md"## Control--Afine Systems
md"## Control--Affine Systems

Non--linear Systems of the form:
```math
Expand All @@ -594,7 +594,7 @@ Non--linear Systems of the form:

$\implies$ Non--linear in the state but affine in the input/control.

Control--Afine Systems are common in many mechanical systems.
Control--Affine Systems are common in many mechanical systems.

"

Expand Down