Revision history [back]

The bullet manifest says:

• "To eliminate ambiguity about which API is active the btQuaternion constructor from Euler angles is deprecated. Please explicitly construct it and populate it seperately from euler angles, otherwise it is ambiguous as to what convention is being used."

At this point, I would avoid using the bullet types directly, as they will change again in Fuerte. Here's a solution using only the tf typedef:

// translate roll, pitch and yaw into a Quaternion
tf::Quaternion q;
q.setRPY(pcl::deg2rad(data_set[t][6]), 
         pcl::deg2rad(data_set[t][7]), 
         pcl::deg2rad(data_set[t][8]));

Note the reordering from (yaw, pitch, roll) to (roll, pitch, yaw). Blockquote

The For the first question:

ranges.resize((size_t)size);

For the second, the bullet manifest says:

• "To eliminate ambiguity about which API is active the btQuaternion constructor from Euler angles is deprecated. Please explicitly construct it and populate it seperately from euler angles, otherwise it is ambiguous as to what convention is being used."

At this point, I would avoid using the bullet types directly, as they will change again in Fuerte. Here's a solution using only the tf typedef:

// translate roll, pitch and yaw into a Quaternion
tf::Quaternion q;
q.setRPY(pcl::deg2rad(data_set[t][6]), 
         pcl::deg2rad(data_set[t][7]), 
         pcl::deg2rad(data_set[t][8]));

Note the reordering from (yaw, pitch, roll) to (roll, pitch, yaw). Blockquoteyaw).

For the first question:

ranges.resize((size_t)size);
scan.ranges.resize((int)(90/0.25));

For the second, the bullet manifest says:

• "To eliminate ambiguity about which API is active the btQuaternion constructor from Euler angles is deprecated. Please explicitly construct it and populate it seperately from euler angles, otherwise it is ambiguous as to what convention is being used."

At this point, I would avoid using the bullet types directly, as they will change again in Fuerte. Here's a solution using only the tf typedef:

// translate roll, pitch and yaw into a Quaternion
tf::Quaternion q;
q.setRPY(pcl::deg2rad(data_set[t][6]), 
         pcl::deg2rad(data_set[t][7]), 
         pcl::deg2rad(data_set[t][8]));

Note the reordering from (yaw, pitch, roll) to (roll, pitch, yaw).

For Use the first question:std::vector resize method directly:

scan.ranges.resize((int)(90/0.25));

Be careful of that (int) cast. I believe the floating point result will be exact in this example, so truncation should not be an issue. Generally, it is better to round explicitly.

// translate roll, pitch and yaw into a Quaternion
tf::Quaternion q;
q.setRPY(pcl::deg2rad(data_set[t][6]), 
         pcl::deg2rad(data_set[t][7]), 
         pcl::deg2rad(data_set[t][8]));