Data types in CSV

Since 2.7.14 version of Flexmonster, it’s recommended to use a Mapping Object to customize the representation and structure of the CSV data source fields.  See the migration guide from CSV prefixes to the mapping.

Alternatively, you can use special prefixes for column names to indicate how data should be interpreted by Flexmonster Pivot. Note that the Mapping Object provides more options that the approach with prefixes. 

Migrating from CSV prefixes to the Mapping Object

The Mapping Object is a simple and convenient way of defining the fields’ data type, and we recommend using it instead of CSV prefixes.

For easy migration from CSV prefixes to the mapping, see the migration table below:

Supported CSV prefixes

Here is the list of supported prefixes that can be used to customize the CSV data: 

• + – the field is a dimension.
• - – the field is a value.
• m+ – the field is a month. Note that if the field stores month names only (in either short or full form), the field will be recognized by Flexmonster as a field of the "m+" type automatically. If the field contains custom month names, specify its type as "m+" explicitly.
• w+ – the field is a day of the week.
• d+ – the field is a date. Such fields will be split into 3 different fields: Year, Month, and Day. Date formats that are supported by Flexmonster Pivot are described below.
• D+ – the field is a date. You will see these dates as a hierarchy: Year > Month > Day.
• D4+ – the field is a date. You will see these dates as a hierarchy: Year > Quarter > Month > Day.
• ds+ – the field is a date. Such fields will be formatted using a date pattern (default is "dd/MM/yyyy").
• t+ – the field is a time (measure). Such fields will be formatted using "HH:mm:ss" pattern.
• dt+ – the field is a date (measure). Such fields will be formatted using "dd/MM/yyyy HH:mm:ss" pattern.
• id+ – the field is an id of the record.

Here is the minimal CSV data that will treat Year as a dimension, rather than a numeric measure:

Country, +Year, Sales 
US, 2010, 200 
UK, 2010, 100

Supported date formats

To make date column be interpreted as a date, use prefixes d+, D+, and D4+ for CSV columns. Additionally, data from these columns should have a special date format to be understood properly. The pivot table component supports the ISO 8601 date format, for example: "2016-03-20" (just date) or "2016-03-20T14:48:00" (date and time). Other formats aren’t officially supported and may have unexpected results.

Here is an example of CSV data with date columns – Date1 and Date2:

Size, Discount, d+Date1, D+Date2
214 oz, 14, 2009-11-01, 2009-11-09
214 oz, 12, 2010-12-09, 2009-12-09
212 oz, 36, 2009-09-01, 2009-12-01
212 oz, 27, 2009-09-01, 2010-12-02
212 oz, 18, 2010-11-09, 2009-12-11
212 oz, 16, 2009-09-01, 2009-12-20

The pivot table based on this CSV will look as follows:

As you can see, the Date1 column with prefix d+ is split into three separate fields — Year, Month, and Day. In the Field List, the Date1 column will look as follows:

The Date2 column with D+ prefix is interpreted as a hierarchy that can be drilled down to months and days.