Data Access Modes

Auto Mode

This is the default mode for Marvin. It is recommended to leave Marvin in this mode and let it handle everything. When in Auto mode, Marvin will always first try to access data in Local Mode, and if that attempt fails, it automatically switches to Remote Mode .

Local Mode

In Local mode, Tools (marvin.tools) will access data stored locally on your machine. If you specify a file name, then it will open that file. Alternatively, you can provide a MaNGA-ID or a plate-IFU identifier. Marvin will check to see if you have a database (e.g., running Tools (marvin.tools) at Utah) and use that if possible. More likely, you will not have access to the database, so it will look for a FITS file. If neither of those options is successful, then it will download the requested data if automatic downloading is enabled.

Explicit Filename

import marvin
from marvin.tools.cube import Cube
marvin.config.mode = 'local'

# loads a cube from an explicit filepath
cc = Cube(filename='/Users/Brian/mycubes/manga-8485-1901-LOGCUBE.fits.gz')

Local SAS

import marvin
from marvin.tools.cube import Cube
marvin.config.mode = 'local'

# checks for file in your local SAS. If found, loads it.  If not, may download it.
cc = Cube(plateifu='8485-1901')

Remote Mode

In Remote mode, Tools (marvin.tools) will retrieve the data remotely via the API. When dealing with Marvin Tools like Cubes or Spaxels, if a MaNGA-ID or a plate-IFU identifier is provided, a remote API call is made to Marvin running at Utah, where it retrieves the data requested and returns it as a JSON object. Once the data has been acquired, the Marvin Tool object you are dealing with will work as if you had the data locally.

When working in remote mode, data is lazy loaded. This means it only grabs the bare minimum you need. New data are grabbed on demand.

import marvin
from marvin.tools.cube import Cube
marvin.config.mode = 'remote'

# grabs the necessary information from Utah, returns it to instantiate the Cube tool
cc = Cube(plateifu='8485-1901')

Mode Decision Tree


Mode decision tree