Gantt chart using the Timevis package in R for monitoring processes
In the BI processes, we have several jobs that are scheduled to be executed according to a time, or sequence, or any other option that is selected.
So it is necessary to be managed by the BI team which are the processes that are being executed, how long it is taking, which is the heaviest process, which system, which area (or data mart), etc. Anyway, all the possibilities that may happen and that we have at hand some tool to make a quick and right decision. Because some data mart may have an urgent need for execution, we can prioritize some process or even improve performance.
As we are using Power BI in the current projects I’m working on, I have not found a Gantt chart that is easy to understand, manipulate and even to visualize the final result efficiently. So I choose to use R as a solution for this problem.
So, I ended up choosing the visualization of the R Timevis package (documentation available below), and to be honest, it was the best choice for me, besides the package having several ways of manipulating the data in the graph, it also has a very practical interaction.
Using the packages below, I made a query in the database (Oracle) using the function “channel” and I obtained the data from a log table created by us, I transformed the data in R and used Timevis to visualize the data of an easier way.
Packages:
· library(shiny)
· library(timevis)
· library(RODBC)
· library(DBI)
· library(dplyr)
· library(lubridate)
· library(reshape2)
I believe that it is not worth me to show the query I made at the bank, because it was something specific and created by us, but I will show the final result to show how we managed to find the right solution.
I also used the function “navbarPage” to visualize the graphics created in different tabs and make it more organized.
Shiny app from R:
First graph by process:
Second graph by system:
Third graph by date:
