custom generated pointcloud2 is displayed as line on rviz

edit

I am new in ROS, We are developing LiDAR camera and I am trying to interface out LiDAR camera with ROS. I am using ROS 1 noetic on windows 10 I write a python script to convert the depth image generated from out camera to PointCloud2 format and then publish the generated PointCloud2 format.

When I try to display the PiontCloud2 on Rviz , the pointcloud is dislayed as line on screen

here is a single PointCloud2 sample

header: 
  seq: 8
  stamp: 
    secs: 1661780280
    nsecs: 110538482
  frame_id: "LiDAR1"
height: 80
width: 260
fields: 
  - 
    name: "x"
    offset: 0
    datatype: 7
    count: 20800
  - 
    name: "y"
    offset: 0
    datatype: 7
    count: 20800
  - 
    name: "z"
    offset: 0
    datatype: 7
    count: 20800
is_bigendian: True
point_step: 12
row_step: 3120
data: [130, 115, 134, 191, 118, 171, 231, 190, 23, 217, 78, 64, 103 ....]
is_dense: True
---

I corrected the offset to be 4 for Y and 8 for z

I generate the Point Cloud by converting my depth image using camera parameters, applying the following command and then convert it to string

   #cx shift of camera center from sensor center in x direction measured in pixels
   #cy shift of camera center from sensor center in y direction measured in pixels
   #fx=fy focal length in  meters
   #rows no of rows per frame
   #cols no of columns per frame

    c, r = np.meshgrid(np.arange(-cols/2,cols/2), np.arange(-rows/2,rows/2), sparse=True)
    z = np.float32(depth/1000)
    x = np.float32(z * (c - cx) * pixel_width / fx )
    y = np.float32(z * (r - cy) * pixel_higth / fy )
    ptcloud = np.dstack((x, y, z)).tostring()

my question is where's my mistake, is it in the PointCloud2 Format, or data or in Rviz configuration?