# Zipped CSV A **zipped CSV** export is what you reach for when you want to load FormShare data into R, Stata, Python, SQL, or any tool that prefers normalized tables. It is a Zip archive containing **one CSV per table** in the form's repository — exactly the layout described in "[How does FormShare stores my data?](/fundamentals/repositories/how-does-formshare-stores-my-data.md)", without the flattening that the [flat CSV](/data-management/data-products/flat-csv-one-csv-file.md) applies. ## Where you see it On the form details page of a form with a repository, click **Export data** → pick **Zipped CSV** (private or public). Like every product, this generates a [private or public variant](/data-management/data-products/private-vs-public-products.md) depending on what you choose, and lands in the **Products and tasks** tab when the background task finishes. ## What's in the zip At minimum, the archive contains one CSV per **data table** in the form: * `maintable.csv` — the main submission table (one row per submission for the fields outside any repeat). * One CSV per **repeat group**, named after the repeat — `rpt_members.csv`, `rpt_crops.csv`, and so on. Each row is one repeat instance, with a foreign key back to the parent row. Optionally — controlled by toggles when you generate the product — the archive can also contain: * **One CSV per multi-select field** (filename pattern `_msel_.csv`). Each row is one (submission, selected option) pair. Without this, multi-selects are condensed into a single column in the parent CSV. * **One CSV per lookup table** (filename pattern `lkp_.csv`). These are the reference tables that hold option codes and labels for `select_one` and `select_multiple` choices. > Add screenshot of the "Export data" dialog for Zipped CSV with the multiselect-tables and lookup-tables toggles visible. Each CSV's columns are the field names from the repository — the same names you see in the [data dictionary](/data-management/data-dictionary.md) — plus standard FormShare bookkeeping columns: the **submission ID** (`rowuuid`), the **primary key**, and (for repeat tables) the row index inside the repeat. ## Options at generation When you click **Generate**, you pick: * **Lookup / select rendering**: how `select_one` fields appear in their parent CSV. * **Code only** — the column contains the option code (`1`, `m`, …). Most compact, the right pick for analytical work. * **Label only** — the column contains the human-readable label. * **Code and label** — two columns per field, suffixed `_label` for the label. * **Include multi-select tables** — adds a CSV per multi-select. Recommended when you want to filter, count, or join on individual options without parsing joined strings. * **Include lookup tables** — adds a CSV per `select_one` / `select_multiple` lookup. Useful when you want the reference data alongside the submission data, e.g. to build dashboards that need labels and codes both. ## When to use zipped CSV * You're loading data into a statistical tool that expects normalized tables (R `data.frame`, pandas DataFrames, Stata, SQL imports). Each CSV maps to one table; join keys are already in place. * The form has repeats and you want to analyse parent and child levels separately. Flat CSV duplicates parent data per child row, which complicates parent-level analytics; zipped CSV keeps the levels distinct. * The form has many multi-selects and you want to count individual options. Multi-select tables let you `GROUP BY option` without splitting strings. * You expect the file to be consumed by a script. Zipped CSV is easier to parse programmatically than Excel or flat CSV with their formatting quirks. ## When *not* to use zipped CSV * You're handing the file to a non-technical recipient who only knows how to double-click an Excel icon. Give them the [Excel](/data-management/data-products/excel.md) or [flat CSV](/data-management/data-products/flat-csv-one-csv-file.md) instead. * You only need a one-off look at the data. Flat CSV is faster to open and read. * You need the data as a tree of nested objects (one document per submission with sub-arrays for repeats). Use [JSON](/data-management/data-products/json.md) instead. ## Practical tips * **Filenames in the zip reflect the table descriptions in the data dictionary.** If you renamed a repeat's description, the CSV filename uses the new description. * **Encoding is UTF-8 with a header row.** All text fields are quoted; numeric and date fields are not. * **Foreign keys connect tables.** A row in `rpt_members.csv` carries the parent's primary key plus its own row index. A row in `maintable_msel_languages.csv` carries the main table's primary key plus the selected option's code. * **The zip is generated once and cached.** If you generate, download, then submit more data, re-generate to get the new rows. Same as any other product. ## What's next * "[Excel](/data-management/data-products/excel.md)" — a workbook variant for spreadsheet-native consumers. * "[JSON](/data-management/data-products/json.md)" — a tree-structured variant for programmatic consumers. * "[Flat CSV (One CSV file)](/data-management/data-products/flat-csv-one-csv-file.md)" — the denormalized single-file alternative. --- # Agent Instructions: Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter: ``` GET https://docs.formshare.app/data-management/data-products/zipped-csv.md?ask= ``` The question should be specific, self-contained, and written in natural language. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.