Error HTTP 404 not found

[elifrag]

Hello, I am trying to download a subset of data using the cms_download_subset() function but I keep getting the following error: "Failed to collect information from subset-job.
HTTP 404 Not Found."

I tried various datasets along with the one in your example but the error remains.

Can you suggest a fix?
Thank you
Eliza

[pepijn-devries]

Hi Eliza,

CopernicusMarine is transitioning to different services. It would seem that the cms_download_subset() is no longer available. The alternative service uses the ZARR format which is currently not well supported in R (see issue #34).

Unfortunately, I don't have a quick fix on short term. Alternatively, you can use cms_download_stac() to download full data sets. I'll try to report any progress in this area here

[elifrag]

Thank you for your reply.
Much appreciated.

[sermomon]

I also got this error. Please report any progress....

[ruimgbarros]

Hey! I did some workaround using the python version using reticulate. I'm leaving it here in case you want a solution.

library(reticulate)
install_python() 
virtualenv_create(envname = "something")
virtualenv_install("something", packages = c("copernicusmarine"))
use_virtualenv("something", required = TRUE)
py_install("copernicusmarine")

cm <- import("copernicusmarine")

result <- cm$subset(
  dataset_id = "cmems_mod_ibi_phy_anfc_0.027deg-2D_PT1H-m",
  start_datetime = start_date,
  end_datetime = end_date,
  variables = list("thetao"),
  minimum_longitude = -9.64671906997131,
  maximum_longitude = 7.364793639893833,
  minimum_latitude = 36.7143279919028,
  maximum_latitude = 41.884404655537004,
  force_download = TRUE,
)

It's not beautiful. But it will be the code I'll be using. I tried having a look at @pepijn-devries amazing work but I could't simply find a solution without using python this way.

[pepijn-devries]

Hey! I did some workaround using the python version using reticulate. I'm leaving it here in case you want a solution.

library(reticulate)
install_python() 
virtualenv_create(envname = "something")
virtualenv_install("something", packages = c("copernicusmarine"))
use_virtualenv("something", required = TRUE)
py_install("copernicusmarine")

cm <- import("copernicusmarine")

result <- cm$subset(
  dataset_id = "cmems_mod_ibi_phy_anfc_0.027deg-2D_PT1H-m",
  start_datetime = start_date,
  end_datetime = end_date,
  variables = list("thetao"),
  minimum_longitude = -9.64671906997131,
  maximum_longitude = 7.364793639893833,
  minimum_latitude = 36.7143279919028,
  maximum_latitude = 41.884404655537004,
  force_download = TRUE,
)

It's not beautiful. But it will be the code I'll be using. I tried having a look at @pepijn-devries amazing work but I could't simply find a solution without using python this way.

Thanks for sharing this workaround. For the time being I think this is the only way to get this to work in R. I'm still hopeful that this can be fixed in R in the future, without the need of Python. To be continued...

[linnetjessell]

Thanks for the workaround - I am trying to run it this way but getting the following error when it asks me for my Copernicus password:

password:Warning: Password input may be echoed.
Error in py_call_impl(callable, call_args$unnamed, call_args$named) :
RuntimeError: reentrant call inside <_io.BufferedReader name=''>

Can you please suggest a fix?
Thanks!

[pepijn-devries]

@linnetjessell The error you describe here seems to originate from the python package. I suggest you look for answers in that direction.

[ruimgbarros]

Hi @linnetjessell!

I'm sorry, on the code I posted I might have omitted the login process.

I've done it this way:

credentials_file <- os$path$expanduser("~/.copernicusmarine/.copernicusmarine-credentials")
if (os$path$exists(credentials_file)) {
  os$remove(credentials_file)
}

cm$login(user, pwd)

Obviously user and pwd are the login details. What I'm doing there is to check if the file where the credentials should be stored exists and, if it does, deletes them and create a new one. (Theoretically only the last line of code should work, but I think you have already tried).

I hope it helps! :)