Linha Microsiga Protheus - Inglês

Páginas filhas
  • B100001_Smart_Analytics_Customization_Flow

Smart Analytics Customization Flow

Product:

TOTVS Smart Analytics

Versions:

11.80 and 12

Event:

How to customize the project and dashboards of TOTVS Smart Analytics

Step by Step:

Introduction

TOTVS Smart Analytics is a standard project that offers a business intelligence solution that comprises a range of possibilities of views that support the decision taking of end users. By acquiring the product, the customer need to customize it, such as for example, create new reports or add new fields é um projeto padrão que oferece uma solução de inteligência de negócios que abrange uma gama de possibilidades de visões que apoiam a tomada de decisão dos usuários finais. Ao adquirir o produto, o cliente pode sentir a necessidade de realizar customizações, como por exemplo, criar novos relatórios ou incluir novos campos.

This article deals the possible customizations that may be implemented on TOTVS Smart Analytics project. There are 2 different cases of customization, in one of them, the user wishes to customize the default project, in order to add new fields and/or new tables, to create new reports with these new data. In another situation, the user intends to create reports or dashboards, using what already exists on the default project. This last case does not demand modifications on the default project, that is, if that's the case, you will need only to create the reports/dashboards directly on the front-end.

 

Customization of Reports and Dashboards

On Smart Analytics you may customize metrics, reports and dashboards, regardless of the customization of new fields or datasets, previously. On front-end, the following cases of customization may be applied:

 

Project Customization - Modeling and ETL

If the customer needs to add new fields or tables to add more information in its project, you will need to customize it. Customization, generally, interferes in 3 main points of the project: modeling, ETL and the configuration file my.properties of Agent. All of these aspects will be mentioned in this article.

The user that customizes must perform it on a local project, never changing the default project. It is recommended to create a copy of the default project on a local project, where customizations will be deployed.

 

Warning

From the moment that the customer edits a graph or metadata that belongs to the default project, the customized project starts to be a complete responsibility of the user, regarding project maintenances and updates.

 

Below, the 3 possible customization scenarios on Smart Analytics will be displayed, which will be analyzed one by one.

 

    If the customer needs extra fields on a Dataset, whether fact or dimension, you may configure the free fields on Smart Analytics default project. 10 free fields are available for each Dataset, considering that on fact tables, there are 5 free fields that represent metrics, 3 free fields that represent date and 2 that represent attributes. For dimensions, the 10 free fields are of the attribute type. These free fields intend to make the inclusion of extra fields on data extraction available, allowing the customer to add new fields, whether it be an indicator or attribute, which are not attended by the default project. This way, new indicators that add new information and may help on decision taking of customers that use TOTVS Smart Analytics.

    By using free fields, the user will no longer need to deploy any customization on the default project. Free fields offer easiness to its users, solving problems that demand customizations, having success on practicality and quickness on TOTVS Smart Analytics maintenance and update.

    This scenario will range when the 10 free fields are not enough, and the customer needs one or more fields in an existing Dataset on the default project, whether it fact or dimension.

    Modeling

    Create a new field on the Dataset you wish, choosing whether fact or attribute.

    ETL

    Add the new fields on the metadata, regarding the Dataset graph edited on modeling.

    Agent - my.properties

    Add the field on the existing search, regarding the Dataset edited.

    The scenario 3 was subdivided in 3 situations (A, B, C), in order to make explanations easier. All of these scenarios display a situation where what will be customized matches a subject totally different from the existing on the default project and, therefore, a new Dataset needs to be created, whether fact or dimension.

     

     

    The first scenario to be studied is situation A that explains how to create Datasets of the Fact type, and its relationship with one or more existing dimensions on the default project.

     

    Modeling

    Add a new Dataset including, necessarily, fields of the fact type in it, and if necessary, of the attribute type. Then, add a relation between the dimensions wished, and this new fact.

     

    ETL

    Create a new graph for load of the fact created. If more than one fact is created, create a graph per fact. There is no need to edit the graph of the dimension associated.

    Agent - my.properties

    Add the search of the Dataset created.

     

    This Situation B demonstrates the customer wish of creating Datasets of the dimension type, and relates it with one or more facts already existent on the default project.

    Modeling

    Add a new Dataset, including fields of the attribute type, only (thus, the Dataset will be defined as a dimension). Then, add a relation between this new dimension and the fact wished.

              

     

    ETL

    Create a new graph for loading the dimension created. If more than one dimension is created, create a graph per dimension. The fact graph related to this dimension, must be updated, in order to add the new field on the metadata, besides adding the reference on mapping the new field.

    Agent - my.properties

    Add the search of the Dataset created. Besides, you will need to add the new field, on the existing search, regarding the dimension related.

    Warning

    Upon selecting the fields that will be created/edited on the search, the order must be the same of the fields on the metadata created/edited on the previous step.

     

    Finally, the last Situation C displays the customer needs of creating Datasets, whether facts and/or dimensions, without relating it to Datasets of the default project. This way, establishing reports only between Facts and Dimensions created by the customer, without intefering with Datasets of the default project.

    Modeling

    Add new Datasets, whether facts as dimensions, then, add a relation between dimensions and facts created.

     

     

    Warning

    Upon completing any of these steps, you must save and publish the new modeling for GoodData cloud, before proceeding to the next steps. To do so, click Publish Model to Server, located on the Menu Details on the right.

    In this moment, the following screen is displayed informing the alterations that will be performed on the modeling available on the cloud.

    Besides, you may also view and copy the MAQL script that adds fields and/or tables customized. To view it, expand the Advanced arrow and click View MAQL:

     

     

    ETL

    Just create a new graph per Dataset created.

    Agent - my.properties

    Add the searches of all Datasets created.

    Warning

    Upon selecting the fields that will be created/edited on search, the order must be the same of the fields on the metadata created/edited on the previous step.

    Notes:

    Release related:

    Flow: Smart Analytics Update