read_csv operator transforms csv files into Tercen datasets. It can upload an individual CSV file or multiple CSV files in a zipped folder.
CSV files (.csv) are text files which use a comma as the delimiter.
(Tercen also has a 'Delimited text' file importer with settings for both comma and other delimiter types. It has more advanced features for data maipulation but cannot be embedded as a file importer in a workflow.)
| Input projection | . |
|---|---|
documentId |
is the documentId (document can be a single csv file, or a zipped set of csv files) |
| Output relations | . |
|---|---|
filename |
character, the name of the csv file |
* |
All columns from the imported CSV file(s) |
| Property | Description |
|---|---|
Headers |
whether the file starts with a header line (default: true) |
Separator |
value separator, Comma or Tab (default: Comma) |
NA String |
strings to interpret as NA values (default: empty) |
Force |
merge files even if their column names differ (default: false) |
Add namespace |
prefix factor names with the step namespace (default: false) |
Column types |
comma-separated name=type pairs forcing column types, e.g. WellId=string,Barcode=string. Supported types: string, double. Empty means automatic type detection. Use this when a column contains number-like values (well IDs, barcodes...) that must be imported as a string factor: types are forced at read time, so leading zeros and long IDs are preserved. |
All columns as string |
import every column as string, disabling automatic type detection (default: false) |
The operator transforms a csv file into a Tercen table. If the document is a ZIP file containing a set of csv files, the operator extracts the csv files and transforms them into Tercen table.
Without Column types, column types are inferred from the data: a column containing only numbers is imported as double (integers and 64-bit integers are converted to double).