The Studio platform can import CSV (Comma-Separated Values) and TSV (Tab-Separated Values) files, and can export CSV files.

| Loader | Characteristic |
| --- | --- |
| File Extensions | `'.csv'`, `'.tsv'`, `'.dsv'` |
| MIME types | `'text/csv'`, `'text/tab-separated-values'`, `'text/dsv'` |
| File Format | \[CSV\] [https://en.wikipedia.org/wiki/comma-separated_values](https://en.wikipedia.org/wiki/comma-separated_values), \[TSV\] [https://en.wikipedia.org/wiki/tab-separated_values](https://en.wikipedia.org/wiki/tab-separated_values) |

## Format Notes

For consistent results, CSV / TSV files should contain:

- a header row (header row auto detection is performed but may not be correct in all cases).
- one or more columns, where each row represents one geospatial feature (there should be some geospatial information).
- all the values in each column should be of the same data type.

Studio will auto detect column types:

- if geospatial columns are found, layers and filters will automatically be created.
- tables without geospatial columns can also be opened and used for analytics, including creation of custom columns, aggregation and joins with other datasets

Example:

```undefined
id,point_latitude,point_longitude,value,start_time
 a,31.2384,-127.30948,5,2019-08-01 12:00
 b,31.2311,-127.30231,11,2019-08-01 12:05
 c,31.2334,-127.30238,9,2019-08-01 11:55
```

## CSV Column Type Detection

Because CSV files do not contain a schema (a specification of the type of data in each column), the Studio Platform will attempt to detect column data types by parsing a sample of data in each column, using the following rules:

| Type | Column values that will trigger type deduction |
| --- | --- |
| **_`boolean`_** | `True`, `False` |
| **_`date`_** | `2019-01-01` |
| **_`integer`_** | `1`, `2`, `3` |
| **_`real`_** | `-74.158`, `40.832` |
| **_`string`_** | `hello`, `world` |
| **_`timestamp`_** | `2018-09-01 00:00`, `1570306147`, `1570306147000` |
| **_`geometry`_** | Geometries (e.g. Polygons, Lines, Points) can be embedded into CSV as WKT strings<br>`POLYGON ((-74.158 40.835, -74.148 40.830, -74.151 40.832, -74.158 40.835))` or GeoJson geometries<br>`{"type":"Polygon","coordinates":[[[-74.158,40.835],[-74.157,40.839],[-74.148,40.830],[-74.150,40.833],[-74.151,40.832],[-74.158,40.835]]]}` |
| **_`struct`_** | Must be formatted as JSON objects in double quotes. For example, `{"category1": 1, "category2": 2}` must be formatted as `"{\"category1\": 1, \"category2\": 2}"`. |

**Note:** Make sure to clean up values such as `N/A`, `Null`, `\N`. If your column contains mixed type, Studio will treat it as **_`string`_** to be safe.

## Geometry in CSV files

CSV files with geometry columns in the following formats will be recognized by Studio's tools

- [**WKT**](https://dev.mysql.com/doc/refman/5.7/en/gis-data-formats.html#gis-wkt-format)
- **GeoJSON geometry**

WKT example

```undefined
id,geometry
1,"POLYGON((0 0,10 0,10 10,0 10,0 0),(5 5,7 5,7 7,5 7, 5 5))"
```

GeoJSON geometry example:

```undefined
id,geometry
1,"{\"type\":\"Polygon\",\"coordinates\":[[[-74.158491,40.835947],[-74.157914,40.83902]]]}"
```

> A CSV GeoJSON column should contain only the geometry part of a GeoJSON Feature, which includes `type` and `coordinates`. It must be a JSON formatted string, with quotes (`"). CSV needs to be correctly quoted:

- double quotes are interpreted as quotes: "abc" => abc
- two double quotes inside a quoted string will be interpreted as a single quote. """a""" => "a"