After the dataset has been loaded it gets displayed in a table.
At the top of the table, there are multiple buttons to manipulate the dataset as a whole. The buttons are pretty much self explanatory, but will still be shortly covered.
- Clear All Filters - Removes every filter set for each column. Active after at least one filter has been set.
- Remove Incomplete Columns - Removes every column which has at least one empty cell. Active if at least one empty cell exists in the dataset.
- Remove Incomplete Rows - Removes every row which has at least one empty cell. Active if at least one empty cell exists in the dataset.
- Remove Columns - Removes the selected columns. Active after at least one column has been selected by left-clicking on a column header while holding CTRL
- Merge Column - Merges two columns into one. Depending on the filetypes of the columns it will either concatenate each row in the columns, or as to select an aggregation before merging them. Active after at least two columns have been selected by left-clicking on a column header while holding CTRL
- Aggregate - Opens a dialog to select fields by which the table should be aggregated
- Show Original - Drops all changes and reloads the dataset
The table below the buttons is split between a content table, where the raw contents are displayed and the header, where additional information about each column can be found.
The table header displays the name of the column, the type, some column statistics and provides a toggle to activate/deactivate the column. Additionally, it is possible to open a context menu to manipulate the column.
The automatic datatype detection descerns between five datatypes: integer, number, location, date and string. The datatype of a column can be changed to another, as long as the detected datatype is a subset of the others. Following detected datatypes can be changed to other datatypes:
- integer → number, date, string
- number → string
- date → string
- location → string
Saving the changes
In the bottom right, there are buttons for downloading the current state of the dashboard as CSV or JSON. With this it is possible to do quick manipulations of the data and share the manipulated datasets with others. The created JSON file conforms to the expected JSON format for Visualizer. Additionally it is possible to download a configuration file or get a configuration link. This file/link saves the steps from the original dataset to the transformed one. This may help to perform always the same transformations on a dataset, where the original data source should not be permanently changed. Therefore the changes happen only in Visualizer.
To demonstrate how the configuration link works, we will perform some changes on the dataset an generate a link. We will:
- Display only Residential entries - Open the context menu in type. Select the checkbox for Residential
- Change datatype of zip to string - double click the datatype in the string header cell. In the selectbox select string
- Remove state and street - hold CTRL and click on the state and street header cell. Then click on remove columns
The table should now look like this:
To test the saving feature, click on the 'Get Link' button, the URL should be automatically added to the clipboard, and paste the URL into the addressbar of the browser. Hitting enter will automatically jump to the dashboard. For now lets get back to the dataset table by clicking 'Change Fields' and the table will be shown modified as we saved it.