Closed lsetiawan closed 2 months ago
@anantmittal or @IshikaKhandelwal any status updates on this? I'd like to update this soon and point to the repo after the demo on Tuesday.
@lsetiawan @vanitech: What do you think should be the question? I can update this, but I don't understand how good
responses are with and without RAG for questions such as "What is dark matter?"
Can you paste here the before and after for the dark matter question. The magellanic stream question is too niche.
Or we can always go with Niki's gamer question in which the evidence was very clear (also niche but a great example)
Or we can always go with Niki's gamer question in which the evidence was very clear (also niche but a great example)
Since this is a scientific setting, let's stick with some sort of science related question.
One idea that is astronomy related is to ask about caustics a project we recently completed. Get started docs can be found at: https://caustics.readthedocs.io/en/latest/tutorials/BasicIntroduction.html and https://caustics.readthedocs.io/en/latest/tutorials/VisualizeCaustics.html.
You can ask: "What is caustics?"
Caustics is a powerful gravitational lensing simulator that can support users from beginner to highly advanced.
Simple google search would lead you to definition of just caustic.
Makes sense, we can try - What is caustics? Though the example needs to compare with olmo before RAG and not Google search.
Though the example needs to compare with olmo before RAG and not Google search.
Right. I was just thinking that it might just pull the regular definition of caustic. 😅
@anantmittal An email to start signing up for tutorials has been sent out to participants. We'll need this ASAP as people will start signing up on what tutorials they want to attend. We have 10 attendees signed up so far. Thanks.
I just updated the Scipy site to include the current tutorial outline and acknowledgement.
What is dark matter?
Without Rag:
Dark Matter is a theoretical particle or collection of particles that is present in galaxies, along with ordinary baryonic matter such as stars, gas, and dust. It was first introduced to help explain the motion and overall appearance of galaxies without including luminous (emissive) matter like stars and gas. The term "dark" refers not to its color but rather its lack of emission or emission of only invisible radiation such as gravitational or kinetic, making it undetectable through traditional methods like observing photons or emitting light.
The term "dark" in the context of dark matter is a bit misleading, as there is no direct evidence for this hypothetical particle; instead, its existence is inferred from various observational results that can't be explained by ordinary baryonic matter alone. Dark Matter is believed to make up about 85% of all the matter in the observable universe, and it's thought to affect the structure and motion of galaxies due to its gravitational pull on invisible luminous matter such as stars, gas, and dust.
In summary, dark matter is a theoretical particle or collection of particles that has not been directly observed, but it helps explain various observations about the structure and behavior of galaxies by providing an explanation for the missing mass needed to balance out the visible matter we can see with our eyes.
With Rag:
Dark matter is a theoretical framework within astrophysics and cosmology that describes the intrinsic motion of galaxies and their motion within galactic clusters, based on observational data of the cosmic microwave background (CMB) from NASA's Cosmic Background Explorer (COBE). Dark matter is believed to make up approximately 85% of the universe's mass-energy density and about 27% of its total energy content.
However, it does not emit, reflect or absorb any electromagnetic radiation, making it invisible or "dark" in comparison to normal matter that emits and absorbs light. Dark matter particles are assumed to be Weakly Interacting Massive Particles (WIMPs), which interact through the weak nuclear force but not through electromagnetism. This means that they do not emit light, allowing astronomers to infer their existence from their gravitational effects on visible matter.
While there is strong evidence for dark matter based on its astrophysical and cosmological observations, it remains unknown what these particles are made of or the exact nature of their properties. Researchers continue to explore various possible particle candidates that could explain the observed dark matter abundance in our universe.
What is astropy?
Without Rag:
A community-driven effort to build a common platform of astronomical tools and data formats, Astpy (short for "astronomy") is an open-source project that aims to create a unified framework for working with data in astronomy. Built on top of NumPy, SciPy, and other popular scientific libraries, Astropy provides functions, classes, and tools to facilitate and simplify the processing and analysis of astronomical datasets and simulations. This platform offers consistency in data structures, formatting, and interfaces across different packages, making it easier for researchers to work with large-scale observational and theoretical datasets alike. The project serves as a central location for many open-source tools, simplifying access and encouraging collaboration among astronomers and computer scientists.
With Rag:
Astropy is a community-developed open-source project that provides an integrated package for astronomical data analysis in Python. It includes various functionalities, such as time management, coordinate transformations, and data manipulation. Astropy is designed to be compatible with other popular data science tools like NumPy, SciPy, and IPython.
The context provided describes the `astropy.utils` package, which contains various utility functions and classes that are not astronomy-specific but mainly for use within the Astropy project. This package covers topics such as downloading data from URLs, data structures, compatibility with older versions of Python/Numpy, and version intercompatibility functions.
To summarize, Astropy is an open-source astronomical data analysis toolkit built upon existing packages like NumPy, SciPy, and IPython to provide a cohesive experience for users seeking astronomy-related functionality in Python. The `astropy.utils` package contains various utility functions and classes that are not specific to astronomy but instead cater to the Astropy project's requirements and internal compatibility concerns.
@lsetiawan @vanitech Added What is Astropy?
.
Need to update the RAG example in https://cfp.scipy.org/2024/talk/BZGQMC/ to be with the astro stuff that we're working on.