[rllib] Moved clip_action into policy_graph; Clip actions in compute_single_action

[SieversLeon]

Implemented clipping for actions in compute_single_action, follow-up to #4413.

[SieversLeon]

Also, what is the general reason behind clipping not being directly embedded in the graph?

[AmplabJenkins]

Can one of the admins verify this patch?

[AmplabJenkins]

Test FAILed.
Refer to this link for build results (access rights to CI server needed):
https://amplab.cs.berkeley.edu/jenkins//job/Ray-PRB/13193/
Test FAILed.

[ericl]

Looks solid, thanks!

[ericl]

Looks like the linter is failing: https://travis-ci.com/ray-project/ray/jobs/186998035

you can run scripts/format.sh to fix

[AmplabJenkins]

Test FAILed.
Refer to this link for build results (access rights to CI server needed):
https://amplab.cs.berkeley.edu/jenkins//job/Ray-PRB/13199/
Test FAILed.

[AmplabJenkins]

Test FAILed.
Refer to this link for build results (access rights to CI server needed):
https://amplab.cs.berkeley.edu/jenkins//job/Ray-PRB/13302/
Test FAILed.

[AmplabJenkins]

Test PASSed.
Refer to this link for build results (access rights to CI server needed):
https://amplab.cs.berkeley.edu/jenkins//job/Ray-PRB/13303/
Test PASSed.

[AmplabJenkins]

Test FAILed.
Refer to this link for build results (access rights to CI server needed):
https://amplab.cs.berkeley.edu/jenkins//job/Ray-PRB/13323/
Test FAILed.

[ericl]

Test failure looks unrelated.