I don't understand why you don't want to make it possible to precompute the Gram matrix any more when calling lars_path. Can you add a test the demonstrate the unexpected behavior with the old version and that is fixed by your change?

The documentation for RandomizedLasso says that precompute can be True, False, or 'auto'. However, when I tried to explicitly pass False, I got an exception.

What exception? BTW any PR for a fix like this requires adding a non regression test to demonstrate the failure that is being resolved by the fix.

Look at #1854 . it mentions the same thing( I hadn't seen this or I would have referenced this already) .. would you like me to add a test?

+1 for a test. I also don't get why we should ignore precompute. There is certainly a problem but I am not convinced it fixes it for all possibilities of precompute.

I passed precompute as an argument in the test_randomized_lasso method and tested it by passing True,False,None and 'auto'.

Indeed LassoLarsIC also need more tests to check the various possible values for precompute. But passing precompute=True should not set Gram=None but force it's precomputation instead.

make sure your code is pep8

how long is the new test to run?

@ogrisel I agree. precompute=True should force the precomputation. Should I add a condition that passes precompute=none in lars_path when precompute is passed as False in RandomizedLasso?

@agramfort The new test is 25 logical lines ( and 44 physical lines).

Also, should I add a similar test for LassoLarsIC?

@ogrisel @agramfort I think I fixed the problem. The Gram can only be passed as None or auto. And so it can not be equated to precompute.
If it is auto it precomputes the matrix from X and if it's None it ignores.
I added a condition if precompute is set to True or auto, then Gram is passed as auto and the precomputation occurs.
However, if precompute is passed as None or False, then Gram is ignored as it should be.
I just checked it on my machine and it passes the test I wrote for various values of precompute. I will pep8 the code and push it here.

Can you please rebase?

randomized_l1 will be removed. CLosing