Oracle Essbase 9 Implementation Guide- P22 potx
Bạn đang xem bản rút gọn của tài liệu. Xem và tải ngay bản đầy đủ của tài liệu tại đây (781.83 KB, 5 trang )
Building the Essbase Outline
[ 90 ]
In our example, we choose a Microsoft Excel le as our data source.
1. Click on the Open data le option
2. Select the le system and browse to the le (for example: c:\data\market
.xls
) and the data in the excel le is as shown in the following screenshot:
3. Select the above excel le and click on OK. You will see the Data Prep
Editor as shown in the following screenshot:
This material is copyright and is licensed for the sole use by Paul Corcorran on 5th July 2009
8601 ave. p #1, , lubbock, , 79423
Download at Boykma.Com
Chapter 3
[ 91 ]
In this screenshot, the upper screen frame shows the data source le in its original
form. The lower screen frame shows how Essbase will interpret the data before
it gets loaded into your Essbase database.
In the example above, we discussed updating the outline from the input data
source le. In the next chapter, we will talk about how to import data using
the second method which is directly loaded from the RDBMS SQL Data Source.
Step 4: Set the Data Source Properties
While plenty of your data will be entered manually from user interaction, a good
deal of it will likely come from external systems. Before Essbase can process any
data le, we need to dene the data to Essbase in the rules le.
Is the data in a comma separated le or tab delimited le? For some feeds, we need
to ignore the header records which are usually the rst row in the input le. These
settings can be set in the Data Source Properties.
Delimiter: Data delimiters are used to separate individual elds in the data source.
The types of delimiters available in Essbase are given as follows:
• Comma: This is for a comma separated data source le.
• Tab: This is for a tab delimited data source le.
• All spaces: The elds will be delimited by empty white spaces.
• Custom: The common delimiter which is agreed with the source systems.
For example, colon (:) or any other single character.
• Column width: The data columns have specic widths. For example,
the rules le will separate the data into columns based on all columns
being ten characters in width.
To open an input data source le in EAS, with the Data Prep Editor open:
1. Click on Options, and then click on the Data Source Properties.
2. The rst tab is the Delimiter tab as shown in the following screenshot:
This material is copyright and is licensed for the sole use by Paul Corcorran on 5th July 2009
8601 ave. p #1, , lubbock, , 79423
Download at Boykma.Com
Building the Essbase Outline
[ 92 ]
3. Select or create the appropriate data delimiter. The appropriate delimiter
is the character or set of characters that are used to separate the columns of
your data. Essbase offers you a few of the most common delimiters and gives
you the option of setting your own delimiters based on what is being used in
your data le.
Step 5: Set the View to Dimension build elds
In this step, we need to dene if this rule is used for Dimension build elds view
or Data load elds view. By default, Data load eld view is selected.
• Dimension build eld: You will select this option when you are using the
rules le with your data le to add new members to the outline or to update
existing members in the outline.
• Data load eld: This option is for loading data with a data load rules le
and will be discussed in more detail later.
To set the Dimension build elds view in EAS, with the Data Prep Editor open:
• Click on View and then select the checkbox Dimension build elds as
we are currently doing a dimension build. We will talk in-depth about
the Data load elds view in the next chapter.
This material is copyright and is licensed for the sole use by Paul Corcorran on 5th July 2009
8601 ave. p #1, , lubbock, , 79423
Download at Boykma.Com
Chapter 3
[ 93 ]
Step 6: Select Dimension build method
Build dimensions methods are basically used for adding, removing, or updating
a member, an alias, or a property of an Essbase outline member. These are dened
in the build rule les.
There are three ways you can build an Essbase outline:
• Generation reference
• Level reference
• Parent-child reference
Let us talk about how each method can be used to update the Essbase outline.
Generation reference
With the generation reference set, the data source should be arranged or organized
from top to bottom as it pertains to the database outline structure. We consider the
dimension name itself as Generation 1. Each record in the rules le contains the
parent name as Generation 1, the child names are Generation 2, children of the
child name are Generation 3, children of that child name are Generation 4, and
so on. For example: Year | Quarter | Month.
This material is copyright and is licensed for the sole use by Paul Corcorran on 5th July 2009
8601 ave. p #1, , lubbock, , 79423
Download at Boykma.Com
Building the Essbase Outline
[ 94 ]
Level reference
With the level reference set, the data source should be arranged or organized from the
leaf node or bottom level to the top level or the dimension name. We consider the leaf
node to be Level 0, the next record is called Level 1, the next record is Level 2, and
so on, until we reach the top dimension level. For example: Month | Quarter | Year.
Parent-child reference
In the parent-child reference, the parent is followed by the child. Each record will
have a parent and then its child. For example, the YEAR member is the parent and
the QUARTER member is the child. The QUARTER level is also a parent since its
child members are the months.
For this example, we have used the generation reference.
GEOGRAPHIC
REGION
REGION
ALIAS
HUB
CITY
CITY
ALIAS
DEALER
NAME
DEALER
ALIAS
NorthAmerica
NorthAmerica
NorthAmerica
NorthAmerica
NorthAmerica
NorthAmerica
NorthAmerica
NorthAmerica
NA
NA
NA
NA
NA
NA
NA
NA
Central
Central
Central
Central
Central
Central
Central
Central
Chicago
Chicago
Chicago
Chicago
Detroit
Detroit
Detroit
Detroit
Motor City EssCar
Motown EssCar
Phord City EssCar
East Side EssCar
Midwest EssCar
Soldier Field EssCar
Capone EssCar
St. Valentines EssCar
12045
12046
12047
12084
17801
17802
17803
17804
DTW
DTW
DTW
DTW
ORD
ORD
ORD
ORD
As shown in the gure above, we can see two things:
• First, we see that the rst record in the data le is a header record, which
usually describes something about the data le itself such as the column
names. This can be skipped by setting 1 in the Number of lines to skip in
the Header section of the Data Source Properties as shown in the screenshot.
Then click on OK.
This material is copyright and is licensed for the sole use by Paul Corcorran on 5th July 2009
8601 ave. p #1, , lubbock, , 79423
Download at Boykma.Com
