Aart/heritability cleanup #112

surgura · 2021-08-09T11:06:27Z

No description provided.

… black whole megacommit

portaloffreedom · 2021-08-09T15:18:10Z

cpprevolve/revolve/gazebo/brains/DifferentialCPGClean.cpp

+    // TODO: Add exception handling
+    sdf::ElementPtr controller_sdf = brain_sdf->GetElement("rv:controller");
+    revolve::DifferentialCPG::ControllerParams params;
+    params.reset_neuron_random = (controller_sdf->GetAttribute("reset_neuron_random")->GetAsString() == "true");


I think we should safe check if all of these SDF parameters are present before we present it to the students :)
Now seems the right moment to do it

I started on this, but the default genome for the examples is directedtree with neural network brain. This already has checks in place. I will add cppn body & brain later after this merge and then I will add the checks to the cpg brain classes.

portaloffreedom · 2021-08-09T15:18:38Z

cpprevolve/revolve/gazebo/brains/DifferentialCPPNCPG.h

+            explicit DifferentialCPPNCPG(const sdf::ElementPtr brain_sdf,
+                                          const std::vector< MotorPtr > &_motors);
+
+				protected:


tabs here, replace with spaces

removed tabs from all files

portaloffreedom · 2021-08-09T15:22:07Z

cpprevolve/revolve/gazebo/motors/ActuatorWrapper.h

+namespace gazebo
+{
+
+class ActuatorWrapper: public revolve::Actuator


double check: is this wrapper still needed or can it be removed from the code?

Use nowhere so removed

portaloffreedom · 2021-08-09T15:53:44Z

cpprevolve/revolve/brains/controller/DifferentialCPG.cpp

@@ -449,7 +552,7 @@ void DifferentialCPG::step(
            // Don't use frame of reference
            else
            {
-                this->output[j] = this->signal_factor_all_*this->abs_output_bound*((2.0)/(1.0 + std::pow(2.718, -2.0*x/this->abs_output_bound)) -1);
+                this->output[j] = this->signal_factor_all_*this->abs_output_bound*((2.0)/(1.0 + std::pow(2.718, -2.0 * x_input / this->abs_output_bound)) - 1);


should we adjust the output to be in between 0 and 1 here? instead of -1;+1

If we do that we need to change in other places as well right? like where it is used and also where input is generated

portaloffreedom · 2021-08-09T15:57:56Z

cpprevolve/revolve/gazebo/motors/MotorFactory.cpp

@@ -58,6 +60,7 @@ MotorPtr MotorFactory::Motor(
 /////////////////////////////////////////////////
 MotorPtr MotorFactory::Create(sdf::ElementPtr _motorSdf)
 {
+  auto coordinates = _motorSdf->GetAttribute("coordinates");


Check for errors in parsing the SDF also here.

portaloffreedom · 2021-08-09T16:12:32Z

pyrevolve/evolution/fitness.py

+    return total_angle
+
+
+# This will not be part of future code, solely for experimental practice


# This will not be part of future code, solely for experimental practice
Then let's remove it XD

pyrevolve/evolution/population/population_management.py

portaloffreedom · 2021-08-09T16:23:09Z

pyrevolve/spec/implementation.py

@@ -441,12 +441,16 @@ def get_epsilon_mutated_parameters(self, params, serialize=False):
        :return: Mutated parameters
        :rtype: dict|list
        """
+


I don't remember what this file is for

I have no idea but its contents appear to be used in revolve

portaloffreedom · 2021-08-09T16:30:31Z

tools/proto2python.sh

@@ -3,7 +3,7 @@
 set -e

 PROTO_FOLDER='../cpprevolve/revolve/gazebo/msgs'
-GAZEBO_PROTO_FOLDER='/home/matteo/Tools/gazebo/include/gazebo-10/gazebo/msgs/proto/'
+GAZEBO_PROTO_FOLDER=:"${HOME}/installed/include/gazebo-10/gazebo/msgs/proto/"


this line is probably wrong, needs to be adjusted for each computer. We could try using pkg-confing to find the correct folder location, or ask the user...

gazebo as submodule?

portaloffreedom · 2021-08-09T16:31:22Z

worlds/plane.recording.world

+                    </camera>
+                    <always_on>1</always_on>
+                    <update_rate>30</update_rate>
+                    <plugin name='RecordingCameraController' filename='libRecorderCameraPlugin.so'>


should we keep this plugin or not XD

it has been exterminated

…sdf parameters. Not yet fully tested but putting it out there so i can continue later

…al function

Aart Stuurman added 2 commits August 9, 2021 13:05

Squash of complete experiment/heritability branch into one superheavy…

5ddea57

… black whole megacommit

Revert .circleci

f988cb3

surgura force-pushed the aart/heritability_cleanup branch from 6d4bbbf to f988cb3 Compare August 9, 2021 11:13

Aart Stuurman added 2 commits August 9, 2021 13:22

Remove check_running_experiments.sh

3f4f41b

Remove camera plugin

00af7e3

surgura force-pushed the aart/heritability_cleanup branch from 2494782 to 00af7e3 Compare August 9, 2021 11:51

Restore original formatting

71cec18

surgura force-pushed the aart/heritability_cleanup branch from c239da1 to 71cec18 Compare August 9, 2021 12:06

Aart Stuurman added 7 commits August 9, 2021 14:29

removed direct_tree_test.py

83df914

Remove brain-speciation experiments

1cd2601

Removed heritability & lsysytem-cppn-cpg experiments

777e82e

Removed more experiments

efc232b

Remove more

8c0a65a

remove revolve.sh

7b48846

Remove start_experiment.sh

f949242

portaloffreedom requested changes Aug 9, 2021

View reviewed changes

Format

3fce2b0

surgura force-pushed the aart/heritability_cleanup branch from ca12125 to 3fce2b0 Compare August 10, 2021 12:29

Aart Stuurman added 7 commits August 10, 2021 15:07

Repair examples/manager_pop.py

68f6cf7

Always add robot_ in sdf builder

e4fabd7

multi_development is now the default everywhere

9183814

Remove limbo and nlopt dependencies. Disabled BO.

3bf6c5f

Fix some wrong function names in examples/manager_pop

0a947e9

Up Cpp version to 17

80f2044

WIP direct tree in manager_pop. Also check previous commit

fb6853e

surgura force-pushed the aart/heritability_cleanup branch from 00ccf6c to fb6853e Compare August 16, 2021 08:02

Aart Stuurman added 2 commits August 16, 2021 10:27

Fix crash in manager_pop

b1e9c00

Add getSdfAttrSafe function util that performs checks before reading …

adf9179

…sdf parameters. Not yet fully tested but putting it out there so i can continue later

surgura force-pushed the aart/heritability_cleanup branch from 90e9c15 to adf9179 Compare August 16, 2021 14:52

Begone, tabs

fb50706

Aart Stuurman added 9 commits August 16, 2021 17:18

Removed ActuatorWrapper

35ed754

Removed dead code

89d0249

fixed typo

b3bcc00

Removed old world

6575593

Added generic body/brain composition. Added CPPN body & brain.

da2a2eb

Rewrote differnetial cpg controller

1662218

Remove unused brain

0470004

I think cppn with targeted brain works?

629b081

fixed cppnneat genotype deserialization function typing to match actu…

9eadb62

…al function

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Aart/heritability cleanup #112

Aart/heritability cleanup #112

surgura commented Aug 9, 2021

portaloffreedom Aug 9, 2021

surgura Aug 16, 2021

portaloffreedom Aug 9, 2021

surgura Aug 16, 2021

portaloffreedom Aug 9, 2021

surgura Aug 16, 2021

portaloffreedom Aug 9, 2021

surgura Aug 16, 2021

portaloffreedom Aug 9, 2021

portaloffreedom Aug 9, 2021

surgura Aug 16, 2021

portaloffreedom Aug 9, 2021

surgura Aug 16, 2021

portaloffreedom Aug 9, 2021

surgura Aug 16, 2021

portaloffreedom Aug 9, 2021

surgura Aug 16, 2021

		return total_angle


		# This will not be part of future code, solely for experimental practice

Aart/heritability cleanup #112

Are you sure you want to change the base?

Aart/heritability cleanup #112

Conversation

surgura commented Aug 9, 2021

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment