Add UserGuide documentation for `IMDReader` #430

amruthesht · 2025-10-10T08:41:41Z

The following PR address Issue #427 and adds UserGuide documentation for the IMDReader via imd.rst and a tutorial/example simulation_imd.ipynb.

The template for the documentation followed is deatiled in the original Issue #427

📚 Documentation preview 📚: https://mdanalysisuserguide--430.org.readthedocs.build/en/430/

review-notebook-app · 2025-10-10T08:41:46Z

Check out this pull request on

See visual diffs & provide feedback on Jupyter Notebooks.

Powered by ReviewNB

Copilot

Pull Request Overview

This PR adds comprehensive UserGuide documentation for the IMDReader class, enabling users to perform real-time streaming analysis of molecular dynamics simulations using the Interactive Molecular Dynamics (IMD) v3 protocol. The documentation follows the template specified in Issue #427.

Adds complete documentation for IMD streaming including setup, usage, and limitations
Provides a comprehensive tutorial notebook with practical examples for different MD engines
Integrates the new documentation into the existing documentation structure

Reviewed Changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated 5 comments.

File	Description
doc/source/formats/reference/imd.rst	Complete IMD format documentation with setup instructions, usage examples, and API reference
doc/source/examples/other/streaming_imd.ipynb	Tutorial notebook demonstrating real-time streaming analysis with practical examples
doc/source/examples/other/README.rst	Updated to include the new streaming_imd tutorial in the documentation index

_{Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.}

doc/source/formats/reference/imd.rst

Copilot · 2025-10-10T16:38:57Z

doc/source/examples/other/streaming_imd.ipynb

+    "### GROMACS Setup\n",
+    "\n",
+    "Add these comprehensive IMD settings to your `.mdp` file:\n",
+    "```code\n",


The code block language identifier should be a valid language (e.g., 'bash', 'text', or 'ini') instead of 'code' for proper syntax highlighting in the documentation.

Copilot · 2025-10-10T16:38:57Z

doc/source/examples/other/streaming_imd.ipynb

+    "### LAMMPS Setup\n",
+    "\n",
+    "Use the comprehensive IMD fix in your input script:\n",
+    "```code\n",


The code block language identifier should be a valid language (e.g., 'bash', 'text', or 'lammps') instead of 'code' for proper syntax highlighting in the documentation.

Suggested change

"```code\n",

"```lammps\n",

Copilot · 2025-10-10T16:38:58Z

doc/source/examples/other/streaming_imd.ipynb

+    "### NAMD Setup\n",
+    "\n",
+    "Add comprehensive IMD configuration to your NAMD configuration file:\n",
+    "```code\n",


The code block language identifier should be a valid language (e.g., 'bash', 'text', or 'tcl') instead of 'code' for proper syntax highlighting in the documentation.

orbeckst · 2025-10-10T16:39:07Z

@ljwoods2 could you also have a quick look please?

ljwoods2 · 2025-10-10T18:00:11Z

doc/source/examples/other/streaming_imd.ipynb

@@ -0,0 +1,325 @@
+{


Line #2. # !pip install imdclient>=0.2.2
Don't we want people to install with Conda for consistency?

Also, will people be running this notebook, or reading it as a guide? I think this cell can potentially be removed, seems like a bit of overkill

Reply via ReviewNB

~~I don't think we need package installation here. imdclient will always be installed when MDA is installed.~~ imdclient will need to be installed separately. However, I'd mention this fact and then link to installation instructions instead of repeating instructions that may change over time.

In general, always link to authoritative instructions; you're the expert on the matter so you should know what the authoritative sources are.

ljwoods2 · 2025-10-10T18:00:11Z

doc/source/examples/other/streaming_imd.ipynb

@@ -0,0 +1,325 @@
+{


Why is IMD-coords essential for GROMACS? Will the stream fail out without it? Or do you just mean its essential for this example?

Maybe it's worth specifying exactly what's necessary for this example: you need to have coordinates turned on regardless of simulation engine, you need to have freq set to 1, and you need to turn off coordinate unwrapping (so that there are consistent configs across engines)

Also, should this link the docs for each of the simulation engines?

Reply via ReviewNB

Link to
MD engine official web pages
imdclient docs https://imdclient.readthedocs.io/en/stable/usage.html

ljwoods2 · 2025-10-10T18:00:12Z

doc/source/examples/other/streaming_imd.ipynb

@@ -0,0 +1,325 @@
+{


Line #42. # simple_imd_analysis()

Why is this commented out? Will github run it automatically as a part of docstring tests if it's uncommented?

Reply via ReviewNB

The notebook needs to be primarily a readable document, don't expect people to edit it. Thus, it should show _working code_, so avoid "un-commenting something so that it works".

If there are issues with it being run, comment on the PR and then let's look for a solution.

ljwoods2 · 2025-10-10T18:00:12Z

doc/source/examples/other/streaming_imd.ipynb

@@ -0,0 +1,325 @@
+{


Line #41. potential_energy = -1000 + np.random.normal(0, 50)
Not all simulation engines implement energy packet, so maybe remove this?

Reply via ReviewNB

That's true but for the ones that do, we should show how one can get the information.

ljwoods2 · 2025-10-10T18:00:12Z

doc/source/examples/other/streaming_imd.ipynb

@@ -0,0 +1,325 @@
+{


Line #51. ax1.clear()
Just double checking, did you test this method of live updating plot locally? Looks a bit different than what we did in workshop

https://github.com/MDAnalysis/imd-workshop-2024/blob/main/activity/graph_utils.py

Reply via ReviewNB

doc/source/formats/reference/imd.rst

ljwoods2 · 2025-10-10T18:06:53Z

doc/source/formats/reference/imd.rst

+
+.. code-block:: bash
+
+   pip install imdclient


I think we encourage conda/mamba installation

Made the required change!

ljwoods2 · 2025-10-10T18:09:00Z

doc/source/formats/reference/imd.rst

+               print(f"WARNING: Large displacement detected at {ts.time} ps: {max_displacement:.2f} Å")
+
+       # Monitor energies if available
+       if 'potential' in ts.data:


I think the key is "potential_energy"

https://github.com/Becksteinlab/imdclient/blob/07069e1ff08d1488936c61e52f922116ecddf210/imdclient/IMDClient.py#L956

For these examples I might not even build in safe guards but just state at the top that you're assuming that energy is transmitted.

ljwoods2 · 2025-10-10T18:14:50Z

doc/source/formats/reference/imd.rst

+
+* ``dt``: Time step size in picoseconds
+* ``step``: Current simulation step number  
+* Energy terms: ``potential``, ``total``, etc. (engine-dependent)


I think these keys are wrong, see:

https://github.com/Becksteinlab/imdclient/blob/07069e1ff08d1488936c61e52f922116ecddf210/imdclient/IMDClient.py#L956

I think we actually merged them into MDA develop docstring for IMDReader incorrectly:

https://imdclient.readthedocs.io/en/latest/protocol_v3.html#energies

Also, are these engine-dependent? It's engine depended whether the energy packet is implemented, but the exact energy keys should be consistent whenever the packet is implemented, right?

@ljwoods2 can you raise an issue, please?

@ljwoods2 - yes, the keys should be the same for all engines, only implementation availability is engine dependent - I chnage the langauage around it

Thanks for pointing out the erros in the API documentation. I'll make a PR to correct it!

ljwoods2 · 2025-10-10T18:22:20Z

doc/source/formats/reference/imd.rst

+           if key not in ['dt', 'step']:
+               print(f"  {key}: {value}")
+
+Multiple Client Connections


There's a ton of edge cases when using multiple clients that we don't test, like when one client tries to change blocking behavior, so I think we should just say it isn't recommended

Discussion here

Becksteinlab/imdclient#114

Maybe should update MDA main codebase docs to say this as well:

https://github.com/amruthesht/mdanalysis/blob/ea85bb510f684ef1f6048ad0a50219665298cbb3/package/MDAnalysis/coordinates/IMD.py#L108

I would leave out the whole section; perhaps mention somewhere that it may be possible but results are not well defined and link to imdclient docs.

@jwoods2 - I actually tested some edge cases out - I'll add it to the relevant PR discussion.

@orbeckst - Done!

ljwoods2 · 2025-10-10T18:23:05Z

doc/source/formats/reference/imd.rst

+
+Most MDAnalysis analysis classes work with streaming data, but some limitations apply:
+
+**Compatible Analysis**


I think this should start by outlining what kinds of analyses will and won't work before jumping into distance and contact analysis

Done - could you check if this is accurate, thanks!

ljwoods2 · 2025-10-10T18:26:32Z

doc/source/formats/reference/imd.rst

+       analysis_data.append(calculate_something(ts))
+
+   # This will NOT work - random access
+   ts = u.trajectory[10]  # TypeError


From reading docstring of __getitem__ of StreamReaderBase, I think these should both be ValueErrors

def __getitem__(self, frame): """Return an iterator for slicing a streaming trajectory. Parameters ---------- frame : slice Slice object. Only the step parameter is meaningful for streams. Returns ------- FrameIteratorAll or StreamFrameIteratorSliced Iterator for the requested slice. Raises ------ TypeError If frame is not a slice object. ValueError If slice contains start or stop values.

If it does throw a TypeError, something is wrong...

Agreed, thanks for pointing it out!

ljwoods2 · 2025-10-10T18:27:15Z

doc/source/formats/reference/imd.rst

+   # This will NOT work - restarting iteration
+   for ts in u.trajectory:
+       break
+   for ts in u.trajectory:  # Won't start from beginning


This suggest it will start from somewhere else other than the beginning, won't it just fail?

Wouldn't it just start from the next available frame in the buffer?

No, it shouldn't if I'm understanding the code right (just trying to run a for loop two times), it should raise an error:

https://github.com/MDAnalysis/mdanalysis/blob/03eef459ddf83e73a776c0e6ddefde4a04ce62c9/testsuite/MDAnalysisTests/coordinates/test_imd.py#L457

tylerjereddy · 2025-10-10T19:08:20Z

doc/source/formats/reference/imd.rst

+       print(f"Time: {ts.time:.2f} ps, Step: {ts.data.get('step', 'N/A')}")
+
+       # Your analysis code here
+       selected_atoms = u.select_atoms("protein and name CA")


to clarify -- is there something about IMD that mandates repeating the selection of the AtomGroup in the loop, or can this be hoisted up for efficiency as usual?

Nothing unique about IMD here, can be changed

@amruthesht move selection out of the loop to model best practices (it's bad for performance!)

@tylerjereddy - nothing specific about IMD requires it. I'll modify the example so it doesn't give the wrong impression

tylerjereddy · 2025-10-10T19:10:08Z

doc/source/formats/reference/imd.rst

+       print(f"Protein COM: {center_of_mass}")
+
+       # Optional: break on some condition
+       if ts.time > 1000:  # Stop after 1000 ps


to clarify: this only stops the analysis, not the running simulation/engine the analysis is connected to?

Since this break is the end of the script, it will trigger the IMDClient to cleanup via atext

https://github.com/Becksteinlab/imdclient/blob/07069e1ff08d1488936c61e52f922116ecddf210/imdclient/IMDClient.py#L146

That cleanup might put the simulation engine in a waiting state for another client connection, but won't kill it.

Whether it makes the simulation engine wait or not depends on 1. how the simulation engine was configured and 2. whether the IMDClient overrode the simulation engine's waiting behavior

continue_after_disconnect kwarg is the IMDClient switch for changing waiting behavior

https://imdclient.readthedocs.io/en/latest/api.html

I wonder if the need to look out for this/think about this might be sensible to mention somewhere inline here then, if only to refer to more detailed docs for someone who may be copy/pasting the sample code as a template.

I'd remove the optional part from this example so that it can stand as a simple example on its own.

Then create another example where you discuss the issue with how breaking the loop may affect the running simulation. This is important but nuanced. If something is complicated, break it into smaller steps, that's basic pedagogy.

Would still recommend breaking up this example

I'd remove the optional part from this example so that it can stand as a simple example on its own.

Then create another example where you discuss the issue with how breaking the loop may affect the running simulation. This is important but nuanced. If something is complicated, break it into smaller steps, that's basic pedagogy.

tylerjereddy · 2025-10-10T19:17:01Z

doc/source/formats/reference/imd.rst

+
+       # Contact analysis
+       selection1 = u.select_atoms("resid 1-10")
+       selection2 = u.select_atoms("resid 50-60")


hoist selections if IMD allows it?

tylerjereddy · 2025-10-10T19:20:44Z

doc/source/formats/reference/imd.rst

+
+* **No timeseries methods**: Cannot use ``trajectory.timeseries()``
+* **No bulk operations**: Cannot extract all data at once
+* **Limited multiprocessing**: Cannot split across processes


Is there a concrete example of something you can do with multiprocessing with "vanilla" MDAnalysis that doesn't work with IMD? I would have figured that if you can get data into a NumPy array you're probably "ok" to split off additional processes if you want to.

By vanilla MDA, does that exclude analysis classes with dask/multiprocessing backend?

You could copy the entire traj into np array as it goes, then create universe with traj from array, but that would defeat the purpose I think

And you can use multiprocessing/parallelism on a single frame / sliding window of frames, I know @HeydenLabASU has tried some windowing and then using parallel algorithms to process an i.e. 200 frame window

I guess maybe this should be changed to say that it only works with the serial analysis backend

that would defeat the purpose I think

I guess you could "buffer"/store a few frames in an array that way without doing IO, and possibly only for a select subset of particles, which does seem within the remit/spirit of what IMD enables, or possibly even one of the more interesting use cases. I think that's what you're saying Matthias is doing.

I might guess that accessing multiple frames in parallel is a problem for the same reason that random access is not available, and that some analyses that require multiple passes and divide work over random-access frame access are also banned, so yeah just trying to think of cases that block multiprocessing usage that aren't already covered by the elementary limitations mentioned. Maybe the multiple clients in different processes thing, but I think you also already ruled them out in comments above, whether in different processes or not.

orbeckst · 2025-10-15T21:19:55Z

doc/source/examples/other/streaming_imd.ipynb

@@ -0,0 +1,325 @@
+{


Include links to the authoritative docs — imdclient and IMDReader in the main docs.

The imdclient package should be installed as a dependency of MDAnalysis (in 2.10.0...IIRC(

Reply via ReviewNB

orbeckst · 2025-10-15T21:19:55Z

doc/source/examples/other/streaming_imd.ipynb

@@ -0,0 +1,325 @@
+{


For all codes:

1) Do we want to send IMD data every integrator step in the example? It's ok to do so but write some text above that says explicitly that we are doing it in the example but in production simulations this may be every 5 or every 10 steps.

For GROMACS:

2) Let's model best practice and encourage everyone to send time and box information so that this is always available in MDAnalysis:

IMD-time = Yes ; Send time information ... IMD-box = Yes ; Send box dimensions (recommended)

Reply via ReviewNB

This section appears to be a duplicate of what's in the imd.rst. Can you just omit all of it and just link?

It's basically impossible to keep documentation in sync so the best you can do is to only have it in a single place and link — DRY, don't repeat yourself.

orbeckst · 2025-10-15T21:19:55Z

doc/source/examples/other/streaming_imd.ipynb

@@ -0,0 +1,325 @@
+{


Explain what analysis you're performing. Link to https://docs.mdanalysis.org/stable/documentation_pages/core/groups.html#MDAnalysis.core.groups.AtomGroup.radius_of_gyration (or somewhere in the User Guide) for background.

Also explain that **as an example** you only process the first 10 frames but normally you'd read all frames.

You should also discuss
skipping (trajectory[::5])
limitations (no start/stop)

Reply via ReviewNB

orbeckst · 2025-10-15T21:19:55Z

doc/source/examples/other/streaming_imd.ipynb

@@ -0,0 +1,325 @@
+{


I don't like the structure of the code in that the contents of the try block are the whole code. If you want to protect code then make a function out of the inner block. However, I would just not use the try/except and just state that the example requires a running simulation.

(Additionally, I don't like catching all exceptions and then just printing an error — this is bad practice because errors can slip through.)

Reply via ReviewNB

orbeckst · 2025-10-15T21:19:55Z

doc/source/examples/other/streaming_imd.ipynb

@@ -0,0 +1,325 @@
+{


Line #21. frame_count = 0
In the text, explain that one of the limitations is that one cannot use trajectory[:10] to make clear why you're keeping track of frames. (You can link to StreamReaderBase for more details on limitations).

By the way, couldn't you use ts.frame? That would at least not be quite as clunky.

Reply via ReviewNB

orbeckst · 2025-10-15T21:19:55Z

doc/source/examples/other/streaming_imd.ipynb

@@ -0,0 +1,325 @@
+{


Line #15. try:
Same comment as above: kind of ugly to have so much code inside try/except.

I'd rewrite the code block inside the try block as a function and then keep the existing try/except structure to demonstrate the overall control structure.

Reply via ReviewNB

orbeckst · 2025-10-15T21:19:55Z

doc/source/examples/other/streaming_imd.ipynb

@@ -0,0 +1,325 @@
+{


Line #41. potential_energy = -1000 + np.random.normal(0, 50)
Why do we fake it? Instead we should get it from the energies packet. You just need to say above that one needs to enable sending of energies and then pull the correct data out of the ts.data. This would be a good example for showing how to do it, don't squander the opportunity.

Reply via ReviewNB

orbeckst · 2025-10-15T21:19:55Z

doc/source/examples/other/streaming_imd.ipynb

@@ -0,0 +1,325 @@
+{


Line #96. # advanced_live_streaming()
No commented out code, see above.

Reply via ReviewNB

orbeckst

Great start for the User Guide. I added a bunch of comments in addition to Tyler's and Lawson's ones.

orbeckst · 2025-10-15T21:23:07Z

doc/source/formats/reference/imd.rst

General comment: link to the primary sources (imdclient docs, IMDReader docs)

orbeckst · 2025-10-15T21:26:02Z

doc/source/formats/reference/imd.rst

+
+.. include:: classes/IMD.txt
+
+Real-time streaming of simulation data between molecular dynamics engines and receiving clients can be achieved using IMD protocols like IMDv2 and IMDv3. The :class:`~MDAnalysis.coordinates.IMD.IMDReader` implements the IMDv3 protocol, enabling live streaming of ongoing simulation data.


You need to explain what the general purpose is (re-summarize paper, IMDReader, imdclient docs). Put yourself in the position of someone whose first contact with the idea is through the User Guide. Answer the questions

What is it?

Why is it useful?

Where can I learn more?

Also explain abbreviation "IMD".

Recyle what you've already have written, e.g. from the "What is streaming" section, which is already great in answering the first question. You just need to taylor the initial 2-3 sentences so that they make sense on their own.

orbeckst · 2025-10-15T21:27:55Z

doc/source/formats/reference/imd.rst

+What is Streaming?
+==================
+
+Streaming involves processing data in real-time as it is generated, rather than storing it for later analysis. In molecular dynamics, this means sending simulation data to a client on-the-fly while the simulation is running, without writing large trajectory files to disk.


This is good!

doc/source/formats/reference/imd.rst

orbeckst · 2025-10-15T21:30:08Z

doc/source/formats/reference/imd.rst

+
+MDAnalysis's :class:`~MDAnalysis.coordinates.IMD.IMDReader` uses the `imdclient <https://imdclient.readthedocs.io/>`_ package and provides a familiar interface for reading streaming data, similar to other trajectory readers in MDAnalysis.
+
+When to Use Streaming?


very nice — answers my questions from the first paragraph: so summarize in the first paragraph at a high level what you're writing here.

orbeckst · 2025-10-16T00:14:17Z

doc/source/formats/reference/imd.rst

+
+Streaming analysis has fundamental constraints due to its real-time nature:
+
+**Data Access Limitations**


use proper section markup for subheadings

orbeckst · 2025-10-16T00:14:20Z

doc/source/formats/reference/imd.rst

+* **No trajectory length**: Total frame count unknown until simulation ends
+* **No independent copies**: Cannot create multiple reader instances for the same stream
+
+**Analysis Constraints**


use proper section markup for subheadings

orbeckst · 2025-10-16T00:15:47Z

doc/source/formats/reference/imd.rst

+* **Limited multiprocessing**: Cannot split across processes
+* **Single client**: Only one reader per IMD stream (engine-dependent)
+
+**Practical Considerations**


use proper section markup for subheadings

orbeckst · 2025-10-16T00:16:50Z

doc/source/formats/reference/imd.rst

+   # This WILL work - forward iteration
+   for ts in u.trajectory:
+       analysis_data.append(calculate_something(ts))
+
+   # This will NOT work - random access
+   ts = u.trajectory[10]  # TypeError


split into separate code blocks with preceding text that states what is being attempted and what will happen; the comments "will work/won't work" can also stay to make it clear.

doc/source/formats/reference/imd.rst

Co-authored-by: ljwoods2 <145226270+ljwoods2@users.noreply.github.com> Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

-Changes to the docs based on comments

doc/source/formats/reference/imd.rst

Co-authored-by: ljwoods2 <145226270+ljwoods2@users.noreply.github.com>

orbeckst

The reST document looks pretty good to me (I think one major comment is unresolved). The notebook still has many unresolved comments as far as I can tell.

orbeckst · 2025-10-30T23:13:49Z

doc/source/formats/reference/imd.rst

+       if ts.time > 1000:  # Stop after 1000 ps
+           break
+
+The ``buffer_size`` parameter (10 MB = 10*1024*1024 bytes in this example) controls the buffer used by `imdclient <https://imdclient.readthedocs.io/>`_ to temporarily store data received from the socket. This buffer accounts for speed differences between the producer (simulation engine) and receiver (analysis code), preventing data loss when analysis is slower than data transmission. A larger buffer is more suitable for systems with many atoms or high transmission frequencies. For more details on buffer management and optimization, see the `imdclient documentation <https://imdclient.readthedocs.io/>`_.


Generally I would try to replace in-line hyperlinks with named links just to keep things tidy, especially if you're using the same URL multiple times. (Optional)

orbeckst · 2025-10-30T23:15:03Z

doc/source/formats/reference/imd.rst

+       print(f"Protein COM: {center_of_mass}")
+
+       # Optional: break on some condition
+       if ts.time > 1000:  # Stop after 1000 ps


Would still recommend breaking up this example

I'd remove the optional part from this example so that it can stand as a simple example on its own.

Then create another example where you discuss the issue with how breaking the loop may affect the running simulation. This is important but nuanced. If something is complicated, break it into smaller steps, that's basic pedagogy.

orbeckst · 2025-10-30T23:30:29Z

RTD installation fails because 3.10 not supported by MDA anymore. (Raised #431 )

orbeckst · 2025-10-30T23:33:38Z

@amruthesht a simple thing to address are the issues from pre-commit https://github.com/MDAnalysis/UserGuide/actions/runs/18957769421/job/54138651607?pr=430

Add: IMD-documentation and tutorial-example

a499695

orbeckst linked an issue Oct 10, 2025 that may be closed by this pull request

Add UserGuide documentation for IMDReader - reading streaming data with IMDv3 protocol #427

Open

orbeckst requested review from Copilot and orbeckst October 10, 2025 16:37

Copilot AI reviewed Oct 10, 2025

View reviewed changes

orbeckst self-assigned this Oct 10, 2025

ljwoods2 reviewed Oct 10, 2025

View reviewed changes

ljwoods2 suggested changes Oct 10, 2025

View reviewed changes

tylerjereddy reviewed Oct 10, 2025

View reviewed changes

orbeckst reviewed Oct 15, 2025

View reviewed changes

orbeckst mentioned this pull request Oct 15, 2025

Release v2.10 MDAnalysis/mdanalysis#5124

Closed

orbeckst requested changes Oct 16, 2025

View reviewed changes

orbeckst and others added 2 commits October 15, 2025 17:23

Apply suggestions from code review

4829f19

Co-authored-by: ljwoods2 <145226270+ljwoods2@users.noreply.github.com> Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

DocUpdate: imd.rst

5ca8a6d

-Changes to the docs based on comments

ljwoods2 suggested changes Oct 17, 2025

View reviewed changes

doc/source/formats/reference/imd.rst Outdated Show resolved Hide resolved

doc/source/formats/reference/imd.rst Outdated Show resolved Hide resolved

Apply suggestions from code review

a644d7d

Co-authored-by: ljwoods2 <145226270+ljwoods2@users.noreply.github.com>

orbeckst requested changes Oct 30, 2025

View reviewed changes

orbeckst mentioned this pull request Oct 30, 2025

RTD build fails when MDA 2.10.0 required #431

Open


		Most MDAnalysis analysis classes work with streaming data, but some limitations apply:

		Compatible Analysis


		.. include:: classes/IMD.txt

		Real-time streaming of simulation data between molecular dynamics engines and receiving clients can be achieved using IMD protocols like IMDv2 and IMDv3. The :class:`~MDAnalysis.coordinates.IMD.IMDReader` implements the IMDv3 protocol, enabling live streaming of ongoing simulation data.


		MDAnalysis's :class:`~MDAnalysis.coordinates.IMD.IMDReader` uses the `imdclient <https://imdclient.readthedocs.io/>`_ package and provides a familiar interface for reading streaming data, similar to other trajectory readers in MDAnalysis.

		When to Use Streaming?


		Streaming analysis has fundamental constraints due to its real-time nature:

		Data Access Limitations

Add UserGuide documentation for IMDReader #430

Are you sure you want to change the base?

Add UserGuide documentation for IMDReader #430

Uh oh!

Conversation

amruthesht commented Oct 10, 2025 • edited by github-actions bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

review-notebook-app bot commented Oct 10, 2025

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull Request Overview

Reviewed Changes

Uh oh!

Uh oh!

Uh oh!

Copilot AI Oct 10, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI Oct 10, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI Oct 10, 2025

Choose a reason for hiding this comment

Uh oh!

orbeckst commented Oct 10, 2025

Uh oh!

ljwoods2 Oct 10, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

orbeckst Oct 15, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

ljwoods2 Oct 10, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

ljwoods2 Oct 10, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

ljwoods2 Oct 10, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

ljwoods2 Oct 10, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

ljwoods2 Oct 10, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Add UserGuide documentation for `IMDReader` #430

Add UserGuide documentation for `IMDReader` #430

amruthesht commented Oct 10, 2025 •

edited by github-actions bot

Loading

ljwoods2 Oct 10, 2025 •

edited

Loading

orbeckst Oct 15, 2025 •

edited

Loading

ljwoods2 Oct 10, 2025 •

edited

Loading

ljwoods2 Oct 10, 2025 •

edited

Loading

ljwoods2 Oct 10, 2025 •

edited

Loading

ljwoods2 Oct 10, 2025 •

edited

Loading

ljwoods2 Oct 10, 2025 •

edited

Loading

ljwoods2 Oct 10, 2025 •

edited

Loading

ljwoods2 Oct 10, 2025 •

edited

Loading