Files on Ouro

Files on Ouro are similar to traditional file storage systems. They can be of any format and are stored as-is, without any additional processing or manipulation. Files are suitable for storing unstructured data, such as documents, images, audio, or video.

Users can upload files of any type, up to 5GB in size.

For many file types, users using the web interface can see rich media visuals of their file, like an image or video. We also support views for a wide range of less common file types:

• Interactive 3D views for supported file formats .glb, .stl, .obj
• Interactive circuit schematic views for .kicad_sch
• Diagrams for Mermaid .mermaid and .mmd files
• Live pages for .html
• Document views for supported file formats .pdf
• Interactive molecule and crystal views for supported file formats .cif, .xyz
• Trajectory playback with energy and force plots for ASE .traj files
• Interactive thermodynamic phase diagrams for .phasediagram, an open format any service can write
• Interactive molecule views for .pdb, .sdf, .mol
• Isosurfaces and slices of volumetric data, such as charge and spin densities, for Gaussian .cube files
• Interactive electronic band structures and densities of states for .bandstructure and .dos, an open format any service can write
• Interactive phonon dispersions and densities of states for .phonons, an open format any service can write

Ouro provides these file views so that the content on the platform is as rich and interactive as it can be. See file formats for how to write each one.

Adding files to Ouro

You can add files to Ouro in a couple ways:

• Using the web interface to upload a file from your computer
• Using the Python SDK

From Python, you can upload files with just a few lines of code.

from ouro import Ouro
 
ouro = Ouro()
file = ouro.files.create(
    name="my_file",
    description="File uploaded from Python",
    visibility="public",
    file_path="path/to/my_file.txt",
)

An agent or app that holds the bytes somewhere Ouro can't read from uses a signed upload instead: it asks for an upload URL, sends the bytes there, and creates the file from the upload.

What Ouro reads from a file

Ouro inspects some files when they are uploaded to decide how to show them. A .cif is recognized as a crystal or a molecule and gets a summary of its structure (formula, space group, lattice); an .xyz is recognized as a single structure or a trajectory. The result is stored on the file as metadata.classification and always reflects the file's current bytes.

Using files

If you find a file useful to your work, you can download it to use. Look for the download button on the file page.

From Python, you can get the URL of a file and download it:

import requests
 
file_id = '9718af43-6562-4485-ae93-81e8f661d496'
file = ouro.files.retrieve(file_id)
file_data = file.read_data()
response = requests.get(file_data.url)
print(response.content)

To pass a file to something that has no Ouro credentials, create a download link. It works for anyone who has it until it expires.

From here, you can do whatever you want with the file.

While raw files can be sufficient for many use cases, there are many scenarios where working with structured data is more beneficial. This is where datasets come in. If you upload a CSV file, the platform will automatically convert it into a dataset. You can then save charts so others see a visualization instead of a raw table. Let's learn more in the next section.