Data sections

Both sample Excel files are translated in a single notebook cell. Unlike most other translate steps, this cell calls a process_file directly rather than a job_file — the translate process file contains two entries, one per table.

Prerequisites

Load dataset metadata must be complete. Sampling logs must exist before samples can be managed.

Notebook cell

In load_ai4sh_sample_data.ipynb, the Translate geotag sample cell runs:

process_file = 'import_data/sample/process/translate_geotag_sample.json'

structured_process_D, scheme_params_D = Initiate_process(notebook_path, scheme_file, process_file)

if structured_process_D is not None:
    Run_process(structured_process_D, scheme_params_D)

Note the cell uses process_file, not job_file — no pilot file is involved.

Translate process file

Path: ./ai4sh/import_data/sample/process/translate_geotag_sample.json

{
  "process": [
    {
      "process": "translate_tabular_data",
      "overwrite": true,
      "parameters": {
        "process": "manage_geolocation",
        "tabular_data_path": "../../sample/excel/AI4SH_geolocation.xlsx",
        "dst_path": "../../sample/manage_process"
      }
    },
    {
      "process": "translate_tabular_data",
      "overwrite": true,
      "parameters": {
        "process": "manage_geolocated_profile_sample",
        "tabular_data_path": "../../sample/excel/AI4SH_samples_geotag.xlsx",
        "dst_path": "../../sample/manage_process"
      }
    }
  ]
}

The two entries are executed in sequence. Each reads an Excel source file and writes one manage process file to ./ai4sh/import_data/sample/manage_process/.

Source files

AI4SH_geolocation.xlsx — one row per sampling point. Required columns:

Column Description
name Unique geolocation identifier (e.g. es_arkaute@10-1)
x_coordinate Longitude (decimal degrees)
y_coordinate Latitude (decimal degrees)

The @ separator in the name is a convention: site@sample_tag. It is not enforced by the framework but is used consistently in the AI4SH project.

AI4SH_samples_geotag.xlsx — one row per sample. Required columns include sampling log name, tag, geolocation name, depth profile bounds, and optional juxtaposition/composition/proximity classifications.

Output

Two manage process files are written:

File Source
manage_geolocation.json AI4SH_geolocation.xlsx
manage_geolocated_profile_sample.json AI4SH_samples_geotag.xlsx

Next step

Proceed to Manage sample geolocation.

Updated: