> For the complete documentation index, see [llms.txt](https://ess-dive.gitbook.io/water-soil-sediment-chemistry-reporting-format/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://ess-dive.gitbook.io/water-soil-sediment-chemistry-reporting-format/instructions/detailed_instructions_data_dictionary_file.md).

# Data dictionary instructions

It is required to include a data dictionary (\_dd.csv) for each csv file in a dataset, or a single data dictionary file for an entire dataset. As mentioned in the detailed instructions for the terminology file, it may be useful to generate a master terminology file that includes the required (and optional) fields for all csv files and then pull from there the relevant fields for each \_dd.csv file. Figure 1 shows an example of how a master terminology file could be structured to comply with the \_dd.csv template as well as the terminology template (i.e., by including the “Unit” and “Column\_or\_Row\_Long\_Name” fields, as well as “Term\_Type”) in order to allow for a single master file collection of all terminology used within a dataset/project/team etc. Note that the Column header “Term” should be replaced by “Column\_or\_Row\_Name” in all \_dd.csv files.

The following instructions are from the [ESS-DIVE file level metadata reporting format](https://github.com/ess-dive-workspace/essdive-file-level-metadata) and are designed to facilitate File-Level Metadata extraction of some fields using the [File Level Metadata Extractor](https://code.ornl.gov/ngee-arctic/ess-dive-meta).

1. Create a CSV data dictionary using the [CSV\_dd\_template](https://github.com/ess-dive-workspace/essdive-file-level-metadata/blob/main/CSV_dd/CSV_dd_template.csv). You can create either:

   a. One data dictionary for each data file or

   b. One data dictionary representing all data files in your dataset
2. See the [CSV\_dd\_quick\_guide](https://github.com/ess-dive-workspace/essdive-file-level-metadata/blob/main/CSV_dd/csv_dd_quick_guide.md) for more details about completing template
3. Enter a new row in the data dictionary for each column/row name in your data matrix

   a. Use "\*" wildcard when the description applies to multiple column/row names

   b. For example - if the same description applies to all spectra data column names in the data file(s)

   1. one row in the data dictionary can refer to "wave\_\*" and this definition will be understood to apply to multiple column names within a data sheet
4. View an example [CSV\_dd\_example](https://github.com/ess-dive-workspace/essdive-file-level-metadata/blob/main/CSV_dd/csv_dd_example.md)
5. Save the CSV\_dd template following the [CSV Reporting Format guidance](https://github.com/ess-dive-workspace/essdive-csv-structure) with the filename "dd.csv" or "\*\_dd.csv". Alternatively you may choose to name you data dictionary:

   a. With the same name as the associated data file, but include '\_dd.csv' at the end of the file name

   b. Create a unique filename for your data dictionary

   c. Incorporate a wildcard into the filename if data dictionary applies to multiple data files (for example - "soil\_cores\_\*\_dd.csv")

### Notes

Following the recommended format and structure of the CSV Reporting Format will facilitate File-Level Metadata extraction of some fields using the [File Level Metadata Extractor](https://code.ornl.gov/ngee-arctic/ess-dive-meta).

![Diagram showing an example terminology file with terms used by a project. This diagram includes terms, their definitions, and other important information about the terms. This template includes both required and optional fields.](https://4286556938-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FS0I71wxBvOCvqINR5a06%2Fuploads%2Fgit-blob-ed0693d0ba13cfec3162959cf63b630711dd59d6%2FData_Dictionary_Figure_1.png?alt=media)

**Figure 1.** Example of a terminology file that is structured to include all terms used by a project (i.e. including headers etc that are required in the \_dd.csv files) - i.e., serving as a complete dictionary. Note that the shaded fields are the only ones required for a terminology file included as part of a water-soil-sed-chem dataset, and the unshaded (appearing white) fields are added because they are required in the \_dd.csv files. If you prefer to keep the data dictionaries and the other terminology files separate, then disregard columns B and D above for the terminology file and remove column E (Term\_Type) for the data dictionary files.