Skip to end of metadata
Go to start of metadata

You are viewing an old version of this page. View the current version.

Compare with Current View Page History

« Previous Version 80 Next »

An indicator drill allows you to further analyse an indicator. Indicator drill consists of an array of blocks. These blocks can be displayed either after clicking on an indicator, or directly on a dashboard.

There are 4 types of blocks:

  • distribution - adjust the distribution of data across the map
  • ranking - order all relevant objects based on the value of selected indicator
  • timeSeries - see the progress and trend of an indicator in a selected time frame
  • categories - list specific classes of data within a single indicator

See the indicator drill blocks in action in the Features Catalog for Developers project.

Syntax

This is an indicator drill which belongs to the Number of Customers indicator from the Retail Solution Demo. It contains all four available block types.

Indicator drill object syntax
{
    "name": "customers_indicator_drill",
    "type": "indicatorDrill",
    "content": {
        "blocks": [
            {
                "type": "distribution"
            },
            {
                "type": "ranking"
            },
            {
                "type": "timeSeries"
            },
            {
                "type": "categories",
                "title": "Customers by channel",
                "description": "Customers by channel",
                "splitProperty": "baskets.on_off_name",
                "orderBy": {
                    "property": "baskets.on_off_name",
                    "direction": "desc"
                }
            },
            {
                "type": "categories",
                "title": "Customers by age and sex",
                "description": "Customers by age and sex categories",
                "splitProperty": "clients.age_name",
                "orderBy": {
                   "property": "clients.age_cat",
                   "direction": "asc"
                },
                "dualProperty": "clients.sex_name",
                "label": "Age categories"
            }
        ]
    }
}
Example of a dumped indicator drill
{
    "url": "/rest/projects/yufqzxkbiecj7jot/md/indicatorDrills/t45dd93ksm4jshrs",
    "dumpTime": "2018-01-31T17:39:08Z",
    "version": "4",
    "content": {
        "id": "t45dd93ksm4jshrs",
        "name": "customers_indicator_drill",
        "type": "indicatorDrill",
        "content": {
            "blocks": [
                {
                    "type": "distribution"
                },
                {
                    "type": "ranking"
                },
                {
                    "type": "timeSeries"
                },
                {
                    "type": "categories",
                    "title": "Customers by channel",
                    "description": "Customers by channel",
                    "splitProperty": "baskets.on_off_name",
                    "orderBy": {
                        "property": "baskets.on_off_name",
                        "direction": "desc"
                    }
                },
                {
                    "type": "categories",
                    "title": "Customers by age and sex",
                    "description": "Customers by age and sex categories",
                    "splitProperty": "clients.age_name",
                    "orderBy": {
                        "property": "clients.age_cat",
                        "direction": "asc",
                        "sort": "asc"
                    },
                    "dualProperty": "clients.sex_name",
                    "label": "Age categories"
                }
            ]
        },
        "accessInfo": {
            "createdAt": "2017-10-05T08:27:04Z",
            "modifiedAt": "2018-01-31T16:39:10Z"
        },
        "links": [
            {
                "rel": "self",
                "href": "/rest/projects/yufqzxkbiecj7jot/md/indicatorDrills/t45dd93ksm4jshrs"
            }
        ]
    }
}

Additional syntax examples

Indicator drill with block displayed on a dashboard
{
    "name": "turnover_value_indicator_drill",
    "type": "indicatorDrill",
    "content": {
        "blocks": [
            {
                "type": "distribution",
                "onDashboard": true
            },
            {
                "type": "ranking"
            },
            {
                "type": "timeSeries"
            }
        ]
    }
}

It is possible to display any block directly on a dashboard by setting the onDashboard property to true (see examples below). The block will be accessible and controllable without having to drill down on the indicator.

Please note that:

  • it is possible to display only one block from each drill
  • in case of the distribution block, histogram filter is displayed instead


Indicator drill with additionalSeries and annotations syntax
{
    "name": "offline_turnover_indicator_drill",
    "type": "indicatorDrill",
    "content": {
        "blocks": [
            {
                "type": "timeSeries",
                "additionalSeries": [
                    {
                        "indicator": "/rest/projects/$projectId/md/indicators?name=online_turnover_indicator"
                    }
                ],
                "annotations": [
                    {
                        "dataset": "/rest/projects/$projectId/md/datasets?name=events"
                    }
                ]
            }
        ]
    }
}

It is possible to extend the timeSeries block with additional time series. In this example, we are comparing the offline and offline turnover, as you can see in the additionalSeries array, which contains a link to the online_turnover_indicator object.

Additionally, it is possible to add annotations. Annotations are basically events that can be displayed in the timeSeries graph. The data for annotations are stored in a specific date subtype dataset, named e.g. "events". This dataset must contain a foreign key property containing id of the object to which the event relates to - in this case shops. It also must be linke to the dim_dates dataset of the date dimension, and contain a description of the event (which also must be in the featureTitle property). The visual representation of such timeSeries block can be found below.

Events dataset syntax
{
    "name": "events",
    "type": "dataset",
    "title": "Events",
    "properties": {
        "featureTitle": {
            "type": "property",
            "value": "description"
        }
    },
    "ref": {
        "type": "dwh",
        "subtype": "date",
        "table": "events",
        "primaryKey": "id",
        "categorizable": false,
        "fullTextIndex": false,
        "properties": [
            {
                "filterable": false,
                "name": "id",
                "title": "Id",
                "column": "id",
                "type": "integer"
            },
            {
                "filterable": false,
                "name": "shop_id",
                "title": "Shop id",
                "column": "shop_id",
                "type": "integer",
                "foreignKey": "shops"
            },
            {
                "filterable": false,
                "name": "date_iso",
                "title": "Date iso",
                "column": "date_iso",
                "type": "date",
                "foreignKey": "dim_dates"
            },
            {
                "filterable": false,
                "name": "description",
                "title": "Description",
                "column": "description",
                "type": "string"
            }
        ]
    }
}
Events dataset data example
id,shop_id,date_iso,description
1,10,2018-07-01,Shortened opening hours
2,10,2018-07-19,Closed for interior redesign
3,10,2018-08-17,Shop opened again

Key description

content

KeyTypeOptionalityDescriptionConstraints
blocksarrayREQUIRED

an array of blocks


content.blocks.distribution

Distribution block's scale and the distribution of the data is defined in its paired indicator.content.

Distribution also has a description, which is not configurable through metadata, but is compiled from the properties of its paired indicator, according to this formula: "Distribution of indicator.title in granularity"

KeyTypeOptionalityDescriptionConstraints
typestring

REQUIRED

type identifier, always distribution

[distribution]
onDashboardbooleanOPTIONALdetermines if the block should be displayed directly on a dashboard[true, false]

content.blocks.ranking

Ranking block can be switched to display either markers, or areas (when they're both available in the view). When there is a higher number of elements, infinite scroll is used.

Ranking also has a description, which is not configurable through metadata, but is compiled from the properties of its paired indicator, according to this formula: "Rank of object according to indicator.title" (note: object in this case is either a polygon from a certain granularity, or a marker).

KeyTypeOptionalityDescriptionConstraints
typestring

REQUIRED

type identifier, always ranking

[ranking]
onDashboardbooleanOPTIONALdetermines if the block should be displayed directly on a dashboard[true, false]
featureTypestringOPTIONAL

specifies which feature type should be ranked by default

default = granularity

[granularity, markers]
directionstringOPTIONAL

specifies the direction of ranked objects

default = desc

[asc, desc]

content.blocks.timeSeries

Time series visualises the progress and trend of the respective indicator in a selected time frame. Additionally, it can be extended with time series of another indicator.

However, its time frame is set by view.content.filterGroup.globalDate filter. The component itself also allows the user to set the granularity of the block (days, weeks, months, quarters or years).

Time series also has a description, which is not configurable through metadata, but is compiled from the properties of its paired indicator, according to this formula: "Progress of indicator.title from view.content.filterGroup.globalDate.startDate to view.content.filterGroup.globalDate.endDate"

KeyTypeOptionalityDescriptionConstraints
typestring

REQUIRED

type identifier, always timeSeries

[timeSeries]
onDashboardbooleanOPTIONALdetermines if the block should be displayed directly on a dashboard[true, false]
defaultPeriodstringOPTIONALspecifies the default period dataset selected by default/rest/projects/$projectId/md/datasets?name={datasetName}
additionalSeriesarrayOPTIONALarray of additionalSeries objects
annotationsarrayOPTIONALarray of annotations objects

content.blocks.timeSeries.additionalSeries

Contains link to an additional time series indicator. See the syntax example above.

KeyTypeOptionalityDescriptionConstraints
indicatorstring

REQUIRED

URL link to the indicator object

/rest/projects/$projectId/md/indicators?name={indicatorName}

content.blocks.timeSeries.annotations

Contains link to a dataset with annotation data. See the syntax example above.

KeyTypeOptionalityDescriptionConstraints
datasetstring

REQUIRED

URL link to the dataset object

/rest/projects/$projectId/md/datasets?name={datasetName}

content.blocks.categories

Categories block list specific classes of data. The data are broke down to a categories defined by a splitProperty. These categories are filterable. The title and description of the categories can be set.

The categories are sorted based on their value, unless you specify custom ordering using orderBy. For example - we want categories with days of the week, sorted correctly. In that case, splitProperty will be the day name, and orderBy.property will be a property with integer value for each day.

This block also offers additional configuration, which allows it to be either condensed, or dual.

KeyTypeOptionalityDescriptionConstraints
typestring

REQUIRED

type identifier, always categories

[categories]
titlestringREQUIREDhuman-readable title of the categories block(*)
descriptionstringREQUIREDdescription of the categories block(*)
splitPropertystringREQUIREDidentifier of a dataset property, by which the classes of data will be split{datasetName}.{datasetProperty}
indicatorstringOPTIONAL

URL link to the indicator object

(info) in case the block should not drill its paired indicator from the dashboard

/rest/projects/$projectId/md/indicators?name={indicatorName}
onDashboardbooleanOPTIONALdetermines if the block should be displayed directly on a dashboard[true, false]
collapsedbooleanOPTIONALdetermines if the block should be collapsed when entering the view[true, false]
visualizedbooleanOPTIONALdetermines if the block should be visualized when entering the view (= if qualitative visualization should be enabled, more here)[true, false]
filterablebooleanOPTIONAL

flag that allows you to turn off the default filterability

if not present, the block is filterable, must be set to false to turn it off

[true, false]
hideNullItemsbooleanOPTIONAL

determines if the block should display categories with null or zero values

default = false

[true, false]
sizeLimitintegerOPTIONALlimit the size of displayed categoriesx => 1
orderByobjectOPTIONAL

specifies ordering of the properties in the block

(info) if orderBy is not specified, the categories are ranked by the indicator value

see content.filterGroup.multiSelect.orderBy


displayValues

arrayOPTIONALspecify values of categories which will be displayed by default (when entering the view)
verticalbooleanOPTIONAL

determines if the block should be vertical

(warning) doesn't work with condensed, dualProperty and label properties

[true, false]
condensedbooleanOPTIONAL

determines if the block should appear as condensed or not

(info) see visual representation below

[true, false]
dualPropertystringOPTIONAL

identifier of a dataset property, by which the data will be dually split

(warning) block with dualProperty is always condensed = true

{datasetName}.{datasetProperty}
labelstringOPTIONAL

determines the default label of condensed categories

(warning) it is not possible to set the label for condensed = false block

(*)

Visual representation

Distribution block

Distribution allows you to adjust the distribution of data across the map.

The data in the distribution block below are evenly distributed, and distinctively separate the areas of interest.

Ranking block

This block ranks areas by a specific indicator. In this case, it's the turnover value in Ward administrative units.

Time series block

Time series block shows progress and trend of a selected indicator over a specific period of time.

Time series block with additionalSeries

This block contains one additional indicator in the additionalSeries. It shows the comparison of offline (blue) and online (black) turnover. See the syntax example above.

If you specify more than one additional indicator, a dropdown menu with all other available indicators is displayed.

Time series block with additionalSeries and annotations

Events specified in the annotations array are related to a specific object - in this case a marker. So, note that to see them - we have to drill down on that specific marker. See the syntax example above.

Categories block

Categories block lists specific classes of data.

This is a comparison of Total turnover by each day of the week.

Categories also offers some additional configuration. It is possible to make the block condensed, or make it dually split (age pyramid). In condensed and dual mode, the detail of each category can be shown by hovering over it.

Please note, that the basic categories configuration is omitted in the snippets below. The splitProperty is set to demography_age_1.age_name, which defines the age categories.

regularcondensed
{
    ...
    "condensed": false
}
{
    ...
    "condensed": true,
    "label": "Age groups"
}

dualvertical
{
    ...
    "condensed": true,
    "dualProperty": "demography_age_1.sex",
    "label": "Age groups"
}
{
    ...
    "vertical": true
}

Blocks on a dashboard

Available by setting the onDashboard property to true.

distributioncategories

rankingtimeSeries


  • No labels