Organizing the data and the results of each step is important for some processings that expect to find inputs in given relatives places on the computer.

The users are encouraged to organize the results of each Exwayz executable as follow for being able to easily use advanced functionnalities and to materialize the structure of the mapping pipeline in a folder architeture.

<aside> 💡 Each Exwayz executable writes specifics files where they are called from, see next tutorials for understanding better the content of each folder described above.

</aside>

**slam_session**
├── **1_loop_closure** 
│   ├── lc_observations.yamlc *# Loop closure observations*
│   └── traj_loop_closure.ply *# Loop closed trajectory*
├── **2_georef**
│   ├── calib_gnss_ref_lidar_euler.txt *# Estimated lever arm (3d vec + 3 angles)*
│   ├── calib_gnss_ref_lidar.txt *# Estimated lever arm (4x4 matrix)*
│   ├── gps_lever_arm.txt  *# Estimated lever arm (3d vector)*
│   ├── optimization.log *# Statistics about the optimization*
│   ├── residuals_calib.ply *# Lever arm optimization residuals*
│   ├── residuals_gps.ply *# Residuals between GNSS and fusion trajectories*
│   ├── residuals_odometry.ply *Residuals between odometry and fusion trajectories*
│   ├── traj_fusion_gps.offset *# SLAM-GNSS fusion trajectory offset*
│   └── traj_fusion_gps.ply *# SLAM-GNSS fusion trajectory*
├── **3_merge**
│   ├── merged_<idx>.las *# Merged point cloud (by chunk)*
│   ├── ...
│   └── merged_<idx>.las
├── **frames**
├── log
│   ├── slam_stats_<yymmdd>-<hhmmss>.log *# Statistics about the SLAM execution*
│   └── time_stats.log *# Statistics about the SLAM execution*
├── traj_gnss.ply *# GNSS trajectory in PLY format*
└── traj_odometry.ply *# Odometry trajectory from the SLAM*

Well done!

You can go now to the next tutorial: 3. Run the SLAM