Oracle Analytics Cloud (OAC) - Using Pipelined Table Functions in Answers (Analysis) Dashboards - Relational Data Sources

July 24, 2020 | 9 minute read
Jay Pearson
Consulting Solutions Architect, Business Intelligence
Text Size 100%:

 Background

* This blog was last tested on OAC 105.8.0-133 (Version 5.8) + ADW 18c *

~

This blog explains how to reference relational data sources using a Pipelined Table Function.

See this blog for instructions on how to reference SOAP/REST API's sources using a pipelined table function.

~

This article walks through the steps to use Pipelined Table Functions in Answers (Analysis) Dashboards.

Pipelined table functions provide the ability to create a dynamic table view of the data on the dashboard. The view can be altered based on what is selected in the prompt.

The blog shows how to create a Meeting Planner Dashboard where the user selects their timezone and the destination timezone for the meeting. A Pipelined Table Function is used to generate a list of suggested suitable meeting times for the two timezones.

The blog is divided into five sections:

1) Create and Test Pipelined Function

2) RPD Configurations for Timezone Prompts

3) Create Timezone Dashboard Prompts

4) RPD Configurations for Pipelined Table

5) Create Analysis and Dashboard

The primary design function of each section is illustrated below:

Additionally, the blog covers other OAC Classic topics such as: Dashboard Prompts, Creating Analysis/Dashboards, Session Variables, Initialization Blocks, RPD Layers, Sort Order Columns, Conditional Formatting, PL/SQL, Dummy Table Joins, and more!

 Create and Test Pipelined Function

Download SQL from here to avoid browser copy/paste errors. 

--Create Type
CREATE TYPE OAC_USER.O_MEETING_PLANNER AS OBJECT
(
SORT_BY INTEGER,
MY_TIME   VARCHAR(50),
DEST_TIME VARCHAR(50)
);

--Create Type as Table of Object
CREATE TYPE OAC_USER.T_MEETING_PLANNER AS TABLE OF OAC_USER.O_MEETING_PLANNER;

--Create Function
CREATE OR REPLACE FUNCTION OAC_USER.F_MEETING_PLANNER (r_MY_ZONE IN VARCHAR, r_DEST_ZONE IN VARCHAR)
RETURN T_MEETING_PLANNER PIPELINED AS
i_MY_TIME VARCHAR(50);
i_DEST_TIME VARCHAR(50);
i_SORT_BY NUMBER;
START_MY_TIME TIMESTAMP := CAST(sysdate || '12:00:00 AM' as timestamp);
START_DEST_TIME TIMESTAMP := FROM_TZ(CAST(FROM_TZ(cast(sysdate || ' 12:00:00 AM' as timestamp),(r_MY_ZONE)) AT TIME ZONE 'UTC' AS TIMESTAMP),'UTC') AT TIME ZONE (r_DEST_ZONE);
    BEGIN
       FOR i in 1..24 loop
       SELECT
       i SORT_BY
       ,TO_CHAR(START_MY_TIME,'hh:mi AM') as MY_TIME
       ,TO_CHAR(START_DEST_TIME,'hh:mi AM') as DEST_TIME
       INTO i_SORT_BY, i_MY_TIME, i_DEST_TIME
       FROM DUAL;
       PIPE ROW (O_MEETING_PLANNER(i,i_MY_TIME,i_DEST_TIME));
       START_MY_TIME := START_MY_TIME + 1/24;
       START_DEST_TIME := START_DEST_TIME + 1/24;
      END LOOP;
      RETURN;
   END F_MEETING_PLANNER;
/

--Test Function
SELECT 
mp.MY_TIME
,mp.DEST_TIME
FROM
TABLE(OAC_USER.F_MEETING_PLANNER('America/Denver','America/New_York')) mp
ORDER BY mp.SORT_BY

--Create Timezone_Names table for Prompt
CREATE TABLE OAC_USER.TIMEZONES
AS
(
SELECT DISTINCT TZNAME
FROM V$TIMEZONE_NAMES
)

 RPD Configurations for Timezone Prompts

a) Import OAC_USER.TIMEZONES into the Physical Layer -> Drag to the Business Model Layer.

b) Join to a dummy table in the Physical & Business Layers if necessary.

c) Uncheck Nullable on the column in the Physical Layer

d) Duplicate TZNAME in the Business Model Layer.

e) Rename TZNAME Columns to "My Timezone" and "Destination Timezone".

f) Drag to Presentation Layer.

g) Create an initialization block and session variable for r_MY_ZONE and r_DEST_ZONE.

 Create Timezone Dashboard Prompts

a) Create a Dashboard Prompt with two Prompts for "My Timezone" and "Destination Time Zone".

The prompts are based on "TIMEZONES"."My Timezone" & "TIMEZONES"."Destination Timezone" using the TIMEZONES table created in Step 1.

b) Uncheck "Enable user to select multiple values". Set a default of UTC for each.

c) Set a Request Variable for each. 

d) Add the Dashboard Prompt to the Dashboard

  

e) Test the Dropdown can return the available timezones.

 RPD Configurations for Pipelined Table

a) From the Physical Layer in the RPD -> Create a New Physical Table.

b) Set the default Initialization String to select from the Pipelined Table Function and pass the Session Variables to it.

SELECT 
mp.MY_TIME
,mp.DEST_TIME
FROM
TABLE(OAC_USER.F_MEETING_PLANNER('VALUEOF(NQ_SESSION.r_MY_ZONE)','VALUEOF(NQ_SESSION.r_DEST_ZONE)')) mp

c) Define Columns.

d) Bring into Business and Presentation Layers and join to dummy table if needed.

e) In the Business Layer sort MY_TIME by the SORT_BY column.

 Create Analysis and Dashboard

a) Add "MEETING_PLANNER"."MY_TIME" and "MEETING_PLANNER"."DEST_TIME" to the Analysis.

b) Add conditional formatting to each column.

Red: 12:00 AM;01:00 AM;02:00 AM;03:00 AM;04:00 AM;05:00 AM;06:00 AM;10:00 PM;11:00 PM

Green: 09:00 AM;10:00 AM;11:00 AM;12:00 PM;01:00 PM;02:00 PM;03:00 PM;04:00 PM;05:00 PM

Yellow: 07:00 AM;08:00 AM;06:00 PM;07:00 PM;08:00 PM;09:00 PM

c) Test Dashboard works as expected.

 Want to Learn More?

Click here for a blog written by Jay Pearson in 2015 on using Pipelined Functions to retrieve data from REST APIs.

Click here for more A-Team Oracle Analytics Cloud (OAC) Blogs.

 Summary

This article walked through the steps to use Pipelined Table Functions in Answers (Analysis) Dashboards.

The blog contains a workshop that illustrates how to create a Meeting Planner Dashboard where the user selects their timezone and the destination timezone for the meeting. In the example, a Pipelined Table Function is used to generate a list of suggested suitable meeting times for the two timezones.

On completion of the blog/workshop, readers will have the knowledge to take the sample and convert it to other use cases that may benefit from a dynamic table populated by dashboard prompts.  

Additionally, the blog covered other OAC Classic topics such as: Dashboard Prompts, Creating Analysis/Dashboards, Session Variables, Initialization Blocks, RPD Layers, Sort Order Columns, Conditional Formatting, PL/SQL, Dummy Table Joins, and more!

Jay Pearson

Consulting Solutions Architect, Business Intelligence


Previous Post

Deploying a Highly Available Windows File Server on OCI

Leo Yuen | 5 min read

Next Post


Setting Up ODI Marketplace Agents in High Availability with ATP

Matthieu Lombard | 9 min read