Bug introduced in 10.0.0 and fixed in 10.1.0.

I'm facing a strange behavior of AssociationMap.

With a fresh Kernel session:

In[1]:= Definition@AssociationMap
Out[1]= Attributes[AssociationMap] = {Protected, ReadProtected}

After loading a package I'm developing, running some test against this package I sometimes end up with:

In[113]:= Definition@AssociationMap
Out[113]= Null

and AssociationMap no more evaluate anything.

I didn't ovveride intentionally AssociationMap, and, because it is Protected, I suppose I cannot have ovverride accidentally.

Any help is appreciated.


After a lot of research I found the problem is related with the following minimal code.


AssociationMap[f, {a, b, c}]

The problem appear when putting the two expression in two different cells evaluating the two cells sequentially in a fresh Mathematica session. The problem doesn't appear if:

  1. the two expression are evaluated at once, using "Evaluate Notebook" or putting the code in a single cell;
  2. or one suppress the Timing@;
  3. or one add another AssociationMap cell before the two and evaluate this new cell before the two others.

All tests done with Mathematica v10.0.1.0 under non-Admin account of Windows v8.1 x64.

Mathematica graphics


Wolfram Research confirmed this is a bug:

Hi Andrea, Thanks a lot for sending us the feedback. This is indeed a serious issue and I've filed an incidence report on that. I've added your information to the report to make sure that you will be informed for any updates on this topic. Many thanks again. Regards,

[...] Wolfram Technology Engineer Wolfram Research, Inc.

  • 1
    $\begingroup$ The bugs tag is for confirmed bugs only, so you should wait for feedback from other users before tagging. $\endgroup$
    – Yves Klett
    Commented Dec 6, 2014 at 13:11
  • 1
    $\begingroup$ +1 for a well researched question. However I cannot reproduce your results. I'm using 10.0.0 on OS X though, so it could be local to your version/OS. $\endgroup$
    – C. E.
    Commented Dec 6, 2014 at 13:23
  • $\begingroup$ Sorry @YvesKlett $\endgroup$
    – unlikely
    Commented Dec 6, 2014 at 13:26
  • $\begingroup$ I tested for 10.0.1 for Microsoft Windows (64-bit) and for Mac. I can't reproduce it. $\endgroup$
    – Murta
    Commented Dec 6, 2014 at 14:06
  • 1
    $\begingroup$ I cannot reproduce this in v10.0.1 on Windows 7 (64-bit). Have you tried reinstalling Mathematica? $\endgroup$ Commented Dec 6, 2014 at 15:46

2 Answers 2


It seems I managed to reproduce the problem (M 10.0.1 or 10.0.2, OS X). It appears only when the suggestions bar is turned on and I wait for it to appear after every single output. This does appear to be a bug.

Workaround: turn off the suggestions bar.

Please do report this problem to support at wolfram.com.

enter image description here

The suggestions bar can be very useful for beginners, but it's been known to cause various problems in the past (many of those problems are already fixed). If you don't use the suggestions bar anyway, consider turning it off. The hint to the solution was that you mentioned that this doesn't happen either when the inputs are in the same cell or when you use Evaluate Notebook. Both prevent the suggestions bar form trying to interpret the output of Timing@<||>.

  • 1
    $\begingroup$ Thanks for investigation. As a workaround, I forced an AssociationMap call at the very beginning of my code. Turning off suggestion bar instead apparently solved several unexpected kernel crashes occurring... $\endgroup$
    – unlikely
    Commented Dec 7, 2014 at 15:18

This problem just started for me. AssociationMap was working last month in an existing notebook, now behaving as per report of OP. It exists in both 10.1 and 10.2 under Win7Pro64 (albeit I have both installed - maybe they share resources?). I had long ago turned off the suggestions bar. I have submitted a bug report to WRI. I have no special start-up (e.g. init.m), or customization config.

I have just identified a workaround: at beginning of Mathematica session, do:

  • $\begingroup$ I flagged this post but I don't agree with myself anymore :P $\endgroup$
    – Kuba
    Commented Oct 13, 2015 at 6:06

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.