How to render raw Lidar scan data to PointCloud2?

edit

Hi, thanks for the response first.

1. I need to re-write in python because i need to fusion with another sensor
2. Already tested but failed in compile (this one of the reason i want to rewrite in python)
3. I mean 3 dimensional array not 3D, like [ [x,x,x,x,x],[y,y,y,y,y],[z,z,z,z,z] ] so i can put the point data cleanly ( if supported, that's great!)

1. You can always look up to the package and understand how they have done it. Later you can mimic it in Python. BTW, a fusion of point clouds in Python could take time.
2. You can share those errors in a new question. But please do not post them in this question.
3. A point cloud is defined as a set of data points in space. In 3 dimensional space, it is a set of XYZ points. If you are interested in the PointCloud2 message definition, please check sensor_msgs/PointCloud2.

1. Yeah, thanks for advice
2. No, i wont post in this question. Better i find the raw data and proccess the data and make it on my own :D
3. Already trying , but in rviz did not showing anything. Can you make an example the data format inside uint8 data[] ? It will very helping me. Like [ x,y,z,x,y,z....] or [x,x,x,y,y,y,z,z,z], which one correct formatting for Pointcloud2.data? I have read the docs and still confused. Many tutorial i read just throwing scan data as params to topic publisher, not how the data look like(in format). Just a numbers in array.

You seem to skip essential readings in this stage which is not recommended. The best I can do is provide references. Quoting from the documentation:

The newly revised ROS point cloud message (currently the de facto standard in PCL) now represents arbitrary n-D (n-dimensional) data. Point values can now be of any primitive data types (int, float, double, etc.), and the message can be specified as 'dense', with height and width values, giving the data a 2D structure, e.g., to correspond to an image of the same region in space. For more information on the rationale behind the new structure, see: PCL_March_2010.pdf and pcl_icra2010.pdf

Big thanks! This is help me out to understanding how properly generating PointCloud2 data.

Glad you made it!

More question about PointCloud2.data: Field data is uint8[] , so maximum value is 255. What if i scaled measured distance by percentage its resulting float data, but the field only accept integer, how to putting data value to showed correctly in rviz ? Little bit confused about pointCloud2.fields with datatype with many numbers datatype but the pointcloud2.data is only accept uint8.

You can push float data to point cloud. I am sharing following snippet;

// reserve the memory at the beginning
cloud.data.resize(total_size);
// do something here such as shown below
// define a 3D vector of float
Eigen::Vector3f xyz;
// initializing with dummy values
xyz << 1.2f, 3.4f, 5.6f; 
// copy x to our point cloud
std::memcpy(&cloud.data[curr[0]], &xyz[0], sizeof(float));
// copy y to our point cloud
std::memcpy(&cloud.data[curr[1]], &xyz[1], sizeof(float));
// copy z to our point cloud
std::memcpy(&cloud.data[curr[2]], &xyz[2], sizeof(float));