Use Case - Tableau Extract script

Use Case - Tableau Extract script

Description

There are currently dashboards which display statistics from the ukrr here:

Data portals

As a way of getting real time data out there in the short term there is a requirement to produce flat files which tableau can use to visualise the data. The way tableau works is essentially to take a series of granular groups, sum and apply filters to them.

For example:

 

centre

adultpaed

dialtplt

country 

variable2

measure

value

variable

1

Exeter

Adult

PD

England

Age

Demography

10

>=85

2

Exeter

Adult

PD

England

Age

Ethnicity

5

Black

3

Exeter

Adult

PD

England

Age

Demography

10

75-85

Another way of thinking about it might be a labelled union of all the slices of all the possible pie charts that the stats team might want to make with the data. Tableau then filters and sums the cohort depending on the visualisation.

What is required is a script to produce these visualisations by combining and extending the cohorts produced by the existing calculators and ensuring the existing calculators support all the functionality required to do this.

Target User(s)

Stats team (Shalini and James )

Components/Dependencies

UKRDC database

  • The code mapping table will need to be updated with all the required mappings. Things like mapping centre to region and defining age groups.

Registry codes

  • The vehicle to achieve the above

UKRDC stats dependencies

  • demographics class - the cohort produced by this class should be inclusive of all the patients in the krt calculator. It needs to support all the labelling of the data required by the stats cohort. For example it needs to be able to label a cohort with the region (even though this will be slightly repetitive)

  • krt class - forms the base cohort but shouldn’t require much modification

  • A script in the scripts file that achieves this task from a set of input parameters

Steps

Create xy model
Write core query to …
Write abc function to filter patient cohort …

Reviewers