I noticed several example notebooks (particularly, linking-diagnostics) using a legacy argument to link_df(), namely diagnostics = True. These notebooks import the regular link_df(), but the diagnostics = True keyword has recently been removed. This is not a problem when using the pip version of trackpy (where the diagnostics keyword is still there), but it breaks the examples when using the latest github version of trackpy.
I plan to fix this with existing pull request #44 but I wanted to suggest a "recommended" method for users to import the legacy version of link_df(), in case they do want diagnostics. Here's what I did:
import trackpy as tp
import trackpy.linking.legacy as tplegacy
...
t = tplegacy.link_df(f, 13, memory=3, diagnostics=True)
I noticed several example notebooks (particularly, linking-diagnostics) using a legacy argument to
link_df()
, namelydiagnostics = True
. These notebooks import the regularlink_df()
, but thediagnostics = True
keyword has recently been removed. This is not a problem when using the pip version of trackpy (where the diagnostics keyword is still there), but it breaks the examples when using the latest github version of trackpy.I plan to fix this with existing pull request #44 but I wanted to suggest a "recommended" method for users to import the legacy version of
link_df()
, in case they do want diagnostics. Here's what I did: