Closed abisee closed 4 years ago
Hi @abisee, you can set the data_dir
parameter in download()
to configure the output location. (Check out this part of the docs for more details.)
Great, thanks!
Is there a way to set it once (e.g. setting an environment variable), so I don't need to specify data_dir
every time I use the download()
function?
Unfortunately not, though note that if you've downloaded a Corpus example-corpus
for the first time to a specified data_dir
(aka download('example-corpus', data_dir=[loc])
), future calls to download('example-corpus')
will automatically use the example-corpus
stored at loc
.
However, if you wish to permanently change the download directory, that's not something we support. Typically, we'd imagine most users downloading corpora to a local directory and then loading corpora directly from said local directory (without using the download
function). In other words, using download()
is typically a one-time thing, for when you're first downloading the data to be worked with on your machine.
Would your workflow benefit significantly from allowing the default download directory to be permanently changed? We could include something like this in a future release if we see a common use case for it.
I've tried searching the docs but was unable to find it. Is there something I can do to set the download cache to be something other than
~/.convokit
? Thanks!