scalene VS dtale

I collected a list of profilers (also memory profilers, also specifically for Python) here: https://github.com/albertz/wiki/blob/master/profiling.md

Currently I actually need a Python memory profiler, because I want to figure out whether there is some memory leak in my application (PyTorch based training script), and where exactly (in this case, it's not a problem of GPU memory, but CPU memory).

I tried Scalene (https://github.com/plasma-umass/scalene), which seems to be powerful, but somehow the output it gives me is not useful at all? It doesn't really give me a flamegraph, or a list of the top lines with memory allocations, but instead it gives me a listing of all source code lines, and prints some (very sparse) information on each line. So I need to search through that listing now by hand to find the spots? Maybe I just don't know how to use it properly.

I tried Memray, but first ran into an issue (https://github.com/bloomberg/memray/issues/212), but after using some workaround, it worked now. I get a flamegraph out, but it doesn't really seem accurate? After a while, there don't seem to be any new memory allocations at all anymore, and I don't quite trust that this is correct.

There is also Austin (https://github.com/P403n1x87/austin), which I also wanted to try (have not yet).

Somehow this experience so far was very disappointing.

(Side node, I debugged some very strange memory allocation behavior of Python before, where all local variables were kept around after an exception, even though I made sure there is no reference anymore to the exception object, to the traceback, etc, and I even called frame.clear() for all frames to really clear it. It turns out, frame.f_locals will create another copy of all the local variables, and the exception object and all the locals in the other frame still stay alive until you access frame.f_locals again. At that point, it will sync the f_locals again with the real (fast) locals, and then it can finally free everything. It was quite annoying to find the source of this problem and to find workarounds for it. https://github.com/python/cpython/issues/113939)

Try using scalene to find where your code is running slow (and/or consuming lots of memory). You're talking two different OSs here, there are a ton of things that could explain the difference. But profiling will help you find where the bottleneck is

Use this my fren: https://github.com/plasma-umass/scalene

You should take a look at Scalene - it's even better.

https://github.com/plasma-umass/scalene

I like seeing another Python profiler. The one I've been playing with is Scalene (GitHub). It does some fun things related to letting you see how much things are moving across the system Python memory boundary.

I tried out some features of machine learning models suggesting optimisations on code profiled by scalene and pretty much all of them would make the code less efficient, both time and memory wise. I am not worried. The devil is in the details and ML will not replace all of us anytime soon

Can't say about the floyd-warshall, since that's from elsewhere. However, I'd suggest profiling the code. For example scalene is pretty decent python profiler.

How about scalene?

This is super awesome work and a great technical explanation of a very deep topic.

What happens in the case of JIT or FFI? I think I've only ever seen the Python profiler, scalene[0], handle these cases.

[0]: https://github.com/plasma-umass/scalene

D-Tale: A Python library that provides an interactive web-based interface for data exploration and analysis.

Try this: https://github.com/man-group/dtale

https://github.com/man-group/dtale

I find that I'm actually a lot faster using basic Pandas methods to get the data I want in exactly the form I want it.

If I really want to show everything, I just use:

'''

Here is the list of direct dependencies. I think the majority of the packages I'm using are pretty well-known. Maybe some of the plotly dash packages are aren't as well known (like dash-colorscales) and then some calculation-based packages (squarify, ppscore, missingno) might not be widely used. But as far as I can tell they are harmless. We used D-Tale at my company in an enterprise-style way through jupyterhub.

Hope these help & please support open-source by throwing your star on the repo.

Mostly reliant on pandas since this is a tool specifically designed for pandas. That being said it would be really easy for you to write a simple DB loader to that takes any SQL and returns the results in a pandas dataframe and just pass that to D-Tale. Its actually pretty easy to integrate D-Tale into your own flask/django/streamlit apps. Heres documentation about using it in Flask: https://github.com/man-group/dtale/blob/master/docs/EMBEDDED_FLASK.md

The jupyterhub-server-proxy plugin works great with it. Here's some documentation on how to use it: https://github.com/man-group/dtale#jupyterhub-w-jupyter-server-proxy

These are the most engrossing UI changes I've made in a while so please let me know if you run into any issues. You can play around with them on the demo site (please note that the "Github Fork" link covers the close button for the "sliding side panel" but you can close it using your ESC key). If these changes prove to be easier to use then I can start moving more functionality towards the "sliding side panel" rather than the old popup windows/tabs.

Your data is stored in memory so the size of your dataframe is limited to the memory of your machine. That being said we’ve allowed users to swap out the machanism which stores the data so you can use something like Redis or Shelve to allieviate memory. Here’s some documentation: https://github.com/man-group/dtale/blob/master/docs/GLOBAL_STATE.md