Swarm-SLAM  1.0.0
C-SLAM Framework
Public Member Functions | Public Attributes | List of all members
cslam.algebraic_connectivity_maximization.AlgebraicConnectivityMaximization Class Reference
Inheritance diagram for cslam.algebraic_connectivity_maximization.AlgebraicConnectivityMaximization:
[legend]

Public Member Functions

def __init__ (self, robot_id=0, max_nb_robots=1, max_iters=20, fixed_weight=1.0, extra_params={ "frontend.enable_sparsification":True, "evaluation.enable_sparsification_comparison":False, })
 
def edge_key (self, edge)
 
def replace_weight (self, edge, weight)
 
def update_nb_poses (self, edge)
 
def update_initial_fixed_edge_exists (self, fixed_edge)
 
def set_graph (self, fixed_edges, candidate_edges)
 
def add_fixed_edge (self, edge)
 
def add_candidate_edge (self, edge)
 
def remove_candidate_edges (self, edges, failed=False)
 
def candidate_edges_to_fixed (self, edges)
 
def greedy_initialization (self, nb_candidates_to_choose, edges)
 
def pseudo_greedy_initialization (self, nb_candidates_to_choose, nb_random, edges)
 
def random_initialization (self, nb_candidates_to_choose, edges)
 
def connection_biased_greedy_selection (self, nb_candidates_to_choose, edges, is_robot_included)
 
def compute_offsets (self, is_robot_included)
 
def rekey_edges (self, edges, is_robot_included)
 
def get_included_edges (self, edges, is_robot_included)
 
def fill_odometry (self)
 
def recover_inter_robot_edges (self, edges, is_robot_included)
 
def check_graph_disconnections (self, is_other_robot_considered)
 
def check_initial_fixed_measurements_exists (self, is_robot_included)
 
def run_mac_solver (self, fixed_edges, candidate_edges, w_init, nb_candidates_to_choose)
 
def select_candidates (self, nb_candidates_to_choose, is_other_robot_considered, greedy_initialization=True)
 
def sparsification_comparison_logs (self, rekeyed_candidate_edges, is_robot_included, greedy_result, mac_result)
 
def add_match (self, match)
 

Public Attributes

 fixed_weight
 
 params
 
 fixed_edges
 
 candidate_edges
 
 already_considered_matches
 
 max_iters
 
 max_nb_robots
 
 robot_id
 
 total_nb_poses
 
 initial_fixed_edge_exists
 
 nb_poses
 
 log_greedy_edges
 
 log_mac_edges
 
 offsets
 

Detailed Description

Definition at line 34 of file algebraic_connectivity_maximization.py.

Constructor & Destructor Documentation

◆ __init__()

def cslam.algebraic_connectivity_maximization.AlgebraicConnectivityMaximization.__init__ (   self,
  robot_id = 0,
  max_nb_robots = 1,
  max_iters = 20,
  fixed_weight = 1.0,
  extra_params = {             "frontend.enable_sparsification": True,             "evaluation.enable_sparsification_comparison": False,         } 
) 
Initialization

Args:
    robot_id (int, optional): ID of the robot
    max_nb_robots (int, optional): number of robots. Defaults to 1.
    max_iters (int, optional): maximum number of iterations. Defaults to 20.
    fixed_weight (float, optional): weight of fixed measurements. Defaults to 1.0.

Definition at line 36 of file algebraic_connectivity_maximization.py.

Member Function Documentation

◆ add_candidate_edge()

def cslam.algebraic_connectivity_maximization.AlgebraicConnectivityMaximization.add_candidate_edge (   self,
  edge 
) 
Add a candidate edge to the graph

Args:
    edge (EdgeInterRobot): inter-robot edge

Definition at line 163 of file algebraic_connectivity_maximization.py.

Here is the call graph for this function:
Here is the caller graph for this function:

◆ add_fixed_edge()

def cslam.algebraic_connectivity_maximization.AlgebraicConnectivityMaximization.add_fixed_edge (   self,
  edge 
) 
Add an already computed edge to the graph

Args:
    edge (EdgeInterRobot): inter-robot edge

Definition at line 152 of file algebraic_connectivity_maximization.py.

Here is the call graph for this function:

◆ add_match()

def cslam.algebraic_connectivity_maximization.AlgebraicConnectivityMaximization.add_match (   self,
  match 
) 
Add match

Args:
    match (EdgeInterRobot): potential match

Definition at line 559 of file algebraic_connectivity_maximization.py.

Here is the call graph for this function:

◆ candidate_edges_to_fixed()

def cslam.algebraic_connectivity_maximization.AlgebraicConnectivityMaximization.candidate_edges_to_fixed (   self,
  edges 
) 
Move candidate edges to fixed. 
Use when candidates are successfully converted into measurements

Args:
    edges (list(EdgeInterRobot)): inter-robot edges

Definition at line 192 of file algebraic_connectivity_maximization.py.

Here is the call graph for this function:

◆ check_graph_disconnections()

def cslam.algebraic_connectivity_maximization.AlgebraicConnectivityMaximization.check_graph_disconnections (   self,
  is_other_robot_considered 
) 
Check if the current graph of potential matches is connected

Args:
    is_other_robot_considered: dict(int, bool): indicates which 
                    other robots are are to be included 

Returns:
    dict(int, bool): dict indicating if each robot is connected

Definition at line 391 of file algebraic_connectivity_maximization.py.

Here is the caller graph for this function:

◆ check_initial_fixed_measurements_exists()

def cslam.algebraic_connectivity_maximization.AlgebraicConnectivityMaximization.check_initial_fixed_measurements_exists (   self,
  is_robot_included 
) 
Check if we have an initial fixed measurement with each robot included
If not, greedy selection should be used

Args:
    is_robot_included dict(bool): indicates if each robot is included

Returns:
    bool: check result

Definition at line 419 of file algebraic_connectivity_maximization.py.

Here is the caller graph for this function:

◆ compute_offsets()

def cslam.algebraic_connectivity_maximization.AlgebraicConnectivityMaximization.compute_offsets (   self,
  is_robot_included 
) 
Compute rekey offsets

Args:
    is_robot_included dict(int, bool): Indicates if the robot 
                        is connected and in communication range

Definition at line 291 of file algebraic_connectivity_maximization.py.

Here is the caller graph for this function:

◆ connection_biased_greedy_selection()

def cslam.algebraic_connectivity_maximization.AlgebraicConnectivityMaximization.connection_biased_greedy_selection (   self,
  nb_candidates_to_choose,
  edges,
  is_robot_included 
) 
Greedy weight initialization with connection bias
Prioritize edges that connect unconnected robots.

Definition at line 257 of file algebraic_connectivity_maximization.py.

Here is the call graph for this function:
Here is the caller graph for this function:

◆ edge_key()

def cslam.algebraic_connectivity_maximization.AlgebraicConnectivityMaximization.edge_key (   self,
  edge 
) 
Get a unique key for an edge

Args:
    edge (EdgeInterRobot): inter-robot edge

Returns:
    tuple(int, int, int, int): unique key

Definition at line 76 of file algebraic_connectivity_maximization.py.

Here is the caller graph for this function:

◆ fill_odometry()

def cslam.algebraic_connectivity_maximization.AlgebraicConnectivityMaximization.fill_odometry (   self) 
Add odometry edges
We can infer the odometry edges directly from the number of poses,
without communication.

Returns:
    list(Edge): odometry edges

Definition at line 348 of file algebraic_connectivity_maximization.py.

Here is the caller graph for this function:

◆ get_included_edges()

def cslam.algebraic_connectivity_maximization.AlgebraicConnectivityMaximization.get_included_edges (   self,
  edges,
  is_robot_included 
) 
TODO

Definition at line 337 of file algebraic_connectivity_maximization.py.

Here is the caller graph for this function:

◆ greedy_initialization()

def cslam.algebraic_connectivity_maximization.AlgebraicConnectivityMaximization.greedy_initialization (   self,
  nb_candidates_to_choose,
  edges 
) 
Greedy weight initialization

Args:
    nb_candidates_to_choose (int): number of edges to choose
    edges (list(Edge)): candidate_edges

Definition at line 205 of file algebraic_connectivity_maximization.py.

Here is the caller graph for this function:

◆ pseudo_greedy_initialization()

def cslam.algebraic_connectivity_maximization.AlgebraicConnectivityMaximization.pseudo_greedy_initialization (   self,
  nb_candidates_to_choose,
  nb_random,
  edges 
) 
Greedy weight initialization
    Greedy initialization with for the first nb_candidates_to_choose-nb_random
    then random choice

Args:
    nb_candidates_to_choose (int): number of edges to choose
    nb_random (int): number of edges to choose randomly
    edges (list(Edge)): candidate_edges

Definition at line 220 of file algebraic_connectivity_maximization.py.

Here is the call graph for this function:
Here is the caller graph for this function:

◆ random_initialization()

def cslam.algebraic_connectivity_maximization.AlgebraicConnectivityMaximization.random_initialization (   self,
  nb_candidates_to_choose,
  edges 
) 
Random weight initialization

Args:
    nb_candidates_to_choose (int): number of edges to choose

Definition at line 247 of file algebraic_connectivity_maximization.py.

Here is the call graph for this function:
Here is the caller graph for this function:

◆ recover_inter_robot_edges()

def cslam.algebraic_connectivity_maximization.AlgebraicConnectivityMaximization.recover_inter_robot_edges (   self,
  edges,
  is_robot_included 
) 
Recover IDs from before rekey_edges()

Args:
    edges (list(Edge)): rekeyed edges
    is_robot_included (dict(int, bool)): indicates if a robot is included

Returns:
    list(EdgeInterRobot): edges

Definition at line 364 of file algebraic_connectivity_maximization.py.

Here is the caller graph for this function:

◆ rekey_edges()

def cslam.algebraic_connectivity_maximization.AlgebraicConnectivityMaximization.rekey_edges (   self,
  edges,
  is_robot_included 
) 
Modify keys (nodes ID) from robot_id+keyframe_id to node_id
Result example: 3 robots with 10 nodes eachs
robot 0 nodes id = 1 to 9
robot 1 nodes id = 10 to 19
robot 2 nodes id = 20 to 29

Args:
    edges (dict(EdgeInterRobot)): inter-robot edges
    is_robot_included dict(int, bool): Indicates if the robot 
                        is connected and in communication range

Returns:
    list(Edge): edges with keys for MAC problem

Definition at line 312 of file algebraic_connectivity_maximization.py.

Here is the caller graph for this function:

◆ remove_candidate_edges()

def cslam.algebraic_connectivity_maximization.AlgebraicConnectivityMaximization.remove_candidate_edges (   self,
  edges,
  failed = False 
) 
Remove candidate edge from the graph

Args:
    edges (list(EdgeInterRobot)): inter-robot edges

Definition at line 178 of file algebraic_connectivity_maximization.py.

Here is the call graph for this function:
Here is the caller graph for this function:

◆ replace_weight()

def cslam.algebraic_connectivity_maximization.AlgebraicConnectivityMaximization.replace_weight (   self,
  edge,
  weight 
) 
Replace the weight of an edge

Args:
    edge (EdgeInterRobot): inter-robot edge
    weight (float): new weight

Returns:
    EdgeInterRobot: new edge with new weight

Definition at line 92 of file algebraic_connectivity_maximization.py.

Here is the caller graph for this function:

◆ run_mac_solver()

def cslam.algebraic_connectivity_maximization.AlgebraicConnectivityMaximization.run_mac_solver (   self,
  fixed_edges,
  candidate_edges,
  w_init,
  nb_candidates_to_choose 
) 
Run the maximalization of algebraic connectivity

Args:
    fixed_edges list(EdgeInterRobot): fixed edges
    candidate_edges list(EdgeInterRobot): candidate edges
    w_init (np.array): One-hot selection vector
    nb_candidates_to_choose (int): budger

Definition at line 436 of file algebraic_connectivity_maximization.py.

Here is the call graph for this function:
Here is the caller graph for this function:

◆ select_candidates()

def cslam.algebraic_connectivity_maximization.AlgebraicConnectivityMaximization.select_candidates (   self,
  nb_candidates_to_choose,
  is_other_robot_considered,
  greedy_initialization = True 
) 
Solve algebraic connectivity maximization

Args:
    nb_candidates_to_choose (int): number of candidates to choose,
                    related to a computation/communication budget
    is_other_robot_considered: dict(int, bool): indicates which 
                    other robots are in communication range 
    greedy_initialization: perform greedy initialization based on similarity

Returns:
    list(EdgeInterRobot): selected edges

Definition at line 468 of file algebraic_connectivity_maximization.py.

Here is the call graph for this function:

◆ set_graph()

def cslam.algebraic_connectivity_maximization.AlgebraicConnectivityMaximization.set_graph (   self,
  fixed_edges,
  candidate_edges 
) 
Fill graph struct

Args:
    fixed_edges (list(EdgeInterRobot)): edges that are already computed
    candidate_edges (list(EdgeInterRobot)): candidate edges to compute

Definition at line 132 of file algebraic_connectivity_maximization.py.

Here is the call graph for this function:

◆ sparsification_comparison_logs()

def cslam.algebraic_connectivity_maximization.AlgebraicConnectivityMaximization.sparsification_comparison_logs (   self,
  rekeyed_candidate_edges,
  is_robot_included,
  greedy_result,
  mac_result 
) 
 TODO: document

Definition at line 545 of file algebraic_connectivity_maximization.py.

Here is the call graph for this function:
Here is the caller graph for this function:

◆ update_initial_fixed_edge_exists()

def cslam.algebraic_connectivity_maximization.AlgebraicConnectivityMaximization.update_initial_fixed_edge_exists (   self,
  fixed_edge 
) 
Maintains a bool for each neighbors to know if we have at least a known link.
In cases where no initial fixed edge exists, we perform the greedy selection

Args:
    fixed_edge (EdgeInterRobot): fixed edge in the graph

Definition at line 121 of file algebraic_connectivity_maximization.py.

Here is the caller graph for this function:

◆ update_nb_poses()

def cslam.algebraic_connectivity_maximization.AlgebraicConnectivityMaximization.update_nb_poses (   self,
  edge 
) 
The number of poses should be the maximal edge id known

Args:
    edge (EdgeInterRobot): loop closure edge

Definition at line 110 of file algebraic_connectivity_maximization.py.

Here is the caller graph for this function:

Member Data Documentation

◆ already_considered_matches

cslam.algebraic_connectivity_maximization.AlgebraicConnectivityMaximization.already_considered_matches

Definition at line 59 of file algebraic_connectivity_maximization.py.

◆ candidate_edges

cslam.algebraic_connectivity_maximization.AlgebraicConnectivityMaximization.candidate_edges

Definition at line 58 of file algebraic_connectivity_maximization.py.

◆ fixed_edges

cslam.algebraic_connectivity_maximization.AlgebraicConnectivityMaximization.fixed_edges

Definition at line 57 of file algebraic_connectivity_maximization.py.

◆ fixed_weight

cslam.algebraic_connectivity_maximization.AlgebraicConnectivityMaximization.fixed_weight

Definition at line 54 of file algebraic_connectivity_maximization.py.

◆ initial_fixed_edge_exists

cslam.algebraic_connectivity_maximization.AlgebraicConnectivityMaximization.initial_fixed_edge_exists

Definition at line 67 of file algebraic_connectivity_maximization.py.

◆ log_greedy_edges

cslam.algebraic_connectivity_maximization.AlgebraicConnectivityMaximization.log_greedy_edges

Definition at line 73 of file algebraic_connectivity_maximization.py.

◆ log_mac_edges

cslam.algebraic_connectivity_maximization.AlgebraicConnectivityMaximization.log_mac_edges

Definition at line 74 of file algebraic_connectivity_maximization.py.

◆ max_iters

cslam.algebraic_connectivity_maximization.AlgebraicConnectivityMaximization.max_iters

Definition at line 61 of file algebraic_connectivity_maximization.py.

◆ max_nb_robots

cslam.algebraic_connectivity_maximization.AlgebraicConnectivityMaximization.max_nb_robots

Definition at line 63 of file algebraic_connectivity_maximization.py.

◆ nb_poses

cslam.algebraic_connectivity_maximization.AlgebraicConnectivityMaximization.nb_poses

Definition at line 68 of file algebraic_connectivity_maximization.py.

◆ offsets

cslam.algebraic_connectivity_maximization.AlgebraicConnectivityMaximization.offsets

Definition at line 300 of file algebraic_connectivity_maximization.py.

◆ params

cslam.algebraic_connectivity_maximization.AlgebraicConnectivityMaximization.params

Definition at line 55 of file algebraic_connectivity_maximization.py.

◆ robot_id

cslam.algebraic_connectivity_maximization.AlgebraicConnectivityMaximization.robot_id

Definition at line 64 of file algebraic_connectivity_maximization.py.

◆ total_nb_poses

cslam.algebraic_connectivity_maximization.AlgebraicConnectivityMaximization.total_nb_poses

Definition at line 65 of file algebraic_connectivity_maximization.py.

The documentation for this class was generated from the following file: