Allow squeeze to use negative index

[Kriyszig]

Issue in focus: tensorflow/tfjs#940

Description

squeeze() didn't support the use of negative index which is supported in tensorflow for Python. Added tests for the same in util_test and tensor_test.
The previous implementation didn't allow for random arrangement of values of axis. For example if axes is set to [2, 3], it would remove them both but if you pass it as [3, 2] it would only remove 3 and ignore 2 entirely. I tried the same with tensorflow for python in this Google Colab Notebook and this too was inconsistent with tfjs as tensorflow python allows for removal of axis irrespective of order of input if it's allowed. This inconsistency was removed when adding the support for negative index ( it was totally unintentional but resulted from my implementation to allow negative index ).

Clarification required for following topics:

• I am not sure whether the test messages are satisfactory for the new tests added - Please let me know if this can be improved and I'll implement the same as soon as possible.
• The random input order support was a result of my judgement where i thought [-1, -2] was intuitively much better for a programmer but if we just replaced it with the positive index of same, it would have been [5, 4] which due to it's lack of ascending order would have ignored 4 completely. If this needs to changed, please let me know and I'll come up with another implementation to change this behavior.

Tests Ran:

• yarn test
• yarn lint

BUG

---

For repository owners only:

Please remember to apply all applicable tags to your pull request.
Tags: BUG

For more info see: https://github.com/tensorflow/tfjs/blob/master/DEVELOPMENT.md

---

This change is 

[googlebot]

Thanks for your pull request. It looks like this may be your first contribution to a Google open source project (if not, look below for help). Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA).

📝 Please visit https://cla.developers.google.com/ to sign.

Once you've signed (or fixed any issues), please reply here (e.g. I signed it!) and we'll verify it.

---

What to do if you already signed the CLA

Individual signers

• It's possible we don't have your GitHub username or you're using a different email address on your commit. Check your existing CLA data and verify that your email is set on your git commits.

Corporate signers

• Your company has a Point of Contact who decides which employees are authorized to participate. Ask your POC to be added to the group of authorized contributors. If you don't know who your Point of Contact is, direct the Google project maintainer to go/cla#troubleshoot (Public version).
• The email used to register you as an authorized contributor must be the email used for the Git commit. Check your existing CLA data and verify that your email is set on your git commits.
• The email used to register you as an authorized contributor must also be attached to your GitHub account.

[Kriyszig]

I signed it!

[googlebot]

CLAs look good, thanks!

[Kriyszig]

• Added an exception case in case the given value of axis doesn't lie in [ -shape.length, shape.length) which further makes it consistent with tensorflow for python.
  The behavior of Python in the case can be found in this Google Colab Notebook

Seeking Clarification on following issue:

• The error message was more than 80 characters long and I suppressed TSist using disable-line. Please let me know if this can be replaced with a shorted message, or another approach altogether

Tests Ran:

• yarn test
• yarn lint

[caisq]

Reviewable status: 0 of 1 approvals obtained (waiting on @VisibleMarkov)

---

src/tensor_test.ts, line 1361 at r2 (raw file):

[-1]

Can you also test squeeze([-2, -1])?

---

src/util.ts, line 282 at r2 (raw file):

`Can't squeeze axis ${axis[i]} since its not in [-${shape.length}, ${shape.length}) for shape ${shape}`); // tslint:disable-line

You can break this into two lines by using the string + operator and remove the tslint disable comment.

[Kriyszig]

@caisq Addressed the changes requested in 02d2344
If anything needs to be changed, please let me know and I'll implement it as soon as I can (^_^)

[caisq]

Reviewable status: 0 of 1 approvals obtained (waiting on @VisibleMarkov)

[Kriyszig]

Thanks for the green flag (^_^)
Just curious: Should I wait for another review by someone else?

[caisq]

@VisibleMarkov PR is merged. Thank you for your contribution!