allow DataInterpolations@7

[oscardssmith]

No description provided.

[github-actions]

Benchmark Results (Julia vlts)

Time benchmarks

	master	8e50f18...	master / 8e50f18...
ODEProblem	0.231 ± 0.0027 s	0.239 ± 0.0017 s	0.965 ± 0.013
init	0.0583 ± 0.0016 ms	0.0596 ± 0.0019 ms	0.978 ± 0.041
mtkcompile	0.0405 ± 0.00067 s	0.0436 ± 0.0021 s	0.929 ± 0.047
time_to_load	5.62 ± 0.052 s	5.7 ± 0.089 s	0.984 ± 0.018

Memory benchmarks

	master	8e50f18...	master / 8e50f18...
ODEProblem	1.66 M allocs: 0.0561 GB	1.66 M allocs: 0.0559 GB	1
init	0.908 k allocs: 0.0479 MB	0.908 k allocs: 0.0479 MB	1
mtkcompile	0.256 M allocs: 9.39 MB	0.256 M allocs: 9.39 MB	1
time_to_load	0.153 k allocs: 14.5 kB	0.153 k allocs: 14.5 kB	1

[github-actions]

Benchmark Results (Julia v1)

Time benchmarks

	master	8e50f18...	master / 8e50f18...
ODEProblem	0.232 ± 0.0085 s	0.234 ± 0.012 s	0.992 ± 0.062
init	0.0468 ± 0.011 ms	0.0469 ± 0.012 ms	0.999 ± 0.35
mtkcompile	0.0467 ± 0.00084 s	0.0485 ± 0.0015 s	0.963 ± 0.035
time_to_load	5.82 ± 0.1 s	5.66 ± 0.069 s	1.03 ± 0.022

Memory benchmarks

	master	8e50f18...	master / 8e50f18...
ODEProblem	1.62 M allocs: 0.053 GB	1.62 M allocs: 0.0531 GB	0.999
init	0.818 k allocs: 0.0319 MB	0.818 k allocs: 0.0319 MB	1
mtkcompile	0.282 M allocs: 9.55 MB	0.282 M allocs: 9.55 MB	1
time_to_load	0.159 k allocs: 11.2 kB	0.159 k allocs: 11.2 kB	1

[ChrisRackauckas]

it's just a test dep

[AayushSabharwal]

Tests fail - we're using some deprecated kwargs.

[oscardssmith]

oops. That was a bad race condition. Sorry @AayushSabharwal I'll revert.

[oscardssmith]

that said, why would this cause tests to fail? were we stuck on v6 before?

[AayushSabharwal]

We just didn't update to v7 or v8. Tests fail because they use a kwarg that is removed in later version.

[oscardssmith]

now I'm confused. These tests ran on DataInterpolations v8, so this PR wasn't responsible for failures? Rather than reverting, should we just remove the use of deprecated kwargs?

[AayushSabharwal]

This PR bumped compat but didn't update tests. So tests failed because they used removed functionality. The PR just needs to update tests and up the lower bound.

[oscardssmith]

ah right. Somehow I got it in my head that v8 was already supported prior to this PR, but that's just nonesense. Followup PR incoming