Way 1: Don't use Google Cloud. Just use the existing ESGF infrastructure, which is already pretty great and offers many possibilities for cloud-style data analysis.
-
-
Prikaži ovu nit
-
Way 2: Access the Google Cloud CMIP6 data from you laptop, using
@zarr_dev as a remote data access protocol (similar to OPeNDAP).Prikaži ovu nit -
Way 3: Access the Google Cloud CMIP6 data from a
@ProjectJupyter JupyterHub or Binder running in the same cloud. Things go a bit faster!Prikaži ovu nit -
(All above examples can be reproduced using the binder link in the following repo: https://github.com/pangeo-data/pangeo-cmip6-examples …)
Prikaži ovu nit -
Way 5: Use
@JuliaLanguage to search the Google Cloud CMIP6 archive and load the data with Zarr.jl. Code and binder available athttps://github.com/pangeo-data/pangeo-julia-examples/ …Prikaži ovu nit
Kraj razgovora
Novi razgovor -
-
-
Who pays for CMIP6 storage in the google cloud? Who pays for data downloads? How much does it cost to work on CMIP6 directly in the cloud? How does one pay? What if NSF money not a possibility?Tx.
-
This is part of Google’s public datasets program, so they bear the costs for storage and egress (downloads). Compute cost is the user’s responsibility.
- Još 1 odgovor
Novi razgovor -
-
-
after reading this, committing to learning how to do this over the next few months. can feel myself becoming a
if i don’t keep up! -
The point is not to create more work by introducing more new tools, but to solve very common workflow problems and make us all more productive!
- Još 2 druga odgovora
Novi razgovor -
Čini se da učitavanje traje već neko vrijeme.
Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.