When you connect a new dataset, you may be interest in enhancing your data. For example, you may want to create an artifical column aggregated from other columns or just change a column type. Askdata provides several options to modify and update your data.All the modifications must be done from the user that has the right to access the global data.
From the agent homepage, the first step is to navigate to the dataset panel and then to choose the one you would like to edit.
When the page is loaded you will have an overview of the first 5 rows of your datasets. You can dynamically Remove/add columns just for visualization purposes. You will also have the following menù just above your data:
- Data: An overview of your data
- Schema: Basic information about the columns of your data with the possibility to change/create some features
- Feed: Data Cards linked with this data
- Filter: Specify specific filters allowing/restricting acces to data
- Export: Integration with the Python Askdata SDK
- Relationships: Add a relationship with another existing dataset
- Log: Log of the operations that have been performed
Among all this fields, two need to be deepened: Schema and Filter.
As mentioned earlier, this field provides an overview of the columns of your data. If want to edit a specific column just click on it and you will be prompted with a menù with the following editable options:
- Code*: Code to be used when referring to this column
- Name: Name of the column to be displayed
- Parameter Type: Type of the parameter: Dimension or Measure
- Enabled: Enable the column for queries
- Import Values: Choose if importing values(Yes) or not(NO)
- Description: Description of the column
- Sample Queries: Example of queries to be associated with the column
- Icon: Change the icon of the column. Either pick one or upload yours.
- Mandatory: Set the column as mandatory.
- Synonyms: Synonyms can be manually associated to the column. This means that when I query using the synonyms, the selected column will be displayed.
- Aggregation: Set an aggregation function among AVG, MIN, MAX and SUM
- Indexed With: Specify Technical key for the current column (onyl dimension) to optimize the query
- Searchable: Make elemnts in the results clickable to narrow down the dataset
- Number Formatting: Set a format for the numbers. Various options available
- Number Locale: Set the region you are number refers to: us, en, it or fr.
- Is Date: Yes if this column is a date
- Data Format: Set a custom format for the dates
- Date Formatting: Set a standard format for the dates: dd-MM-yy or dd/MM/yy
- Synonym Generation Logic: You can use a custom logic to generate synonyms with a specific behaviour
- Custom: boolean: Yes if you are creating a new column, No otherwise
- Custom Expression: You can define a custom logic to define a new column
- Ignore Aggregation for Measures: For custom expressions ignores the default aggregation of the measure
- Custom Filter: Inject a custom filter when a specific is requested
- Dynamic Entities: You can define a custom logic to automatically create entities related to that dimension
- Defaul Injections: Set another column to be shown with (i.e. when I query sales also show the year, without specifying the word year). Applied when no other dimensions and measures are requested
- Injections: Set another column to be shown with (i.e. when I query sales also show the year, without specifying the word year). Always applied
- Is Geo Parameter: Yes if this column is a Geo Parameter
- Latitude: Name of the field where to find the value of the latitude
- Longitude: Name of the field where to find the value of the longitude
- External Resource: Path for an external resource where to find the geo attributes
- Key from external resource: The name of the field in the external resource that contains the join field
- Value Formatting: Instructions on how to format the values
Filters are powerful tools that can be used to restrict access to some data. For example if you are a manager in a company and would like one of your employees to access just the data that is related to himself, you could achieve that by defining a simple filter as in the picture:
For example in the above use case we are specifying that if the user login email is firstname.lastname@example.org then just show him the records where name=John and surname=Doe. This obviously is a rather simple example of the definition of a filter, more complex ones can be build in order to guarantee the proper flexibility to your dataset. If you would like to have a deeper overview of filters you can watch this resource.
Once all the edits have been completed, just hit save and the dataset willl be updated.