Tableau - Day 2 (Data Blending Etc)

Download as txt, pdf, or txt
Download as txt, pdf, or txt
You are on page 1of 2

Tableau public vs tableau desktop , tableau reader(optional) (try to install it)

1} limited connections, server side is limited to google sheets, open data etc
2}There will be no live or extract connection (by default it will be only extract)
What you can do is, "Download" the Workbooks from your account which will be saved
as Tableau packaged workbook with extension .twbx and you can share them with
anybody
3}tableau public, you cannot save the workbook on your local machines, only to
tableau public gallery, absolutely no security involved , user level security will
not come into play
4}if somebody has sent you a tableau workbook , you cannot open it using tableau
public, only what you have downloaded from tableau public gallery, you can open
using tableau public else use "tableau reader"
5}Copy function will not be there in tableau public, print to pdf will also not be
there in tableau public

Live , extract-create local database store it in local machine


Live - means direct connection from underlying database
This extract is created by tableau's fast data engine, it is saved with
extension .hyper earlier it was tde(tableau data extract) extract, response is very
qucik with hyper, it will be stored in my local machine
extract-Tableau data extract, simply ecxtract data, you can also extract limited
data also
extract can be copied or exported( image format-.png, data-msaccess format, or
pdf), then save it as twbx (self contained package)
it can be uplodaded onto server (you can publish it on tableau public if data is
generic)

Tableau extract , Tableau live( direct connection to your databse)


this option is present in your tableau desktop
in case of tableau public by default everything is your extract
2 different lines businesses,cant do cross databse connectivuty because my
performance will degrade
join is used -less data less tables
blending-data coming from lot of tables dont want to create one big table , keeping
connections seperate
data blending- two data sources (coffe chain and global superstore)
If we have big databses (like 5 tables from one dataset and 5 tables from another
dataset)
If we can combine at work sheet level with 2 different connections, then efficiency
will improve- Data blending
connection with 2 different databses(global superstore, coffe chain)

The following view is built with help of VIzql server


Rows : profit, category
Columns: region
Joins: Inner join , left join, right join , full outer join
cross database(conncetivity) join (coffee chain)-ms access and im joining with my
sample superstore(excel file)
orange bars, blue bars , tableau is smart enough

Data pane , analytics pane,


column shelf, row shelf, page shelf, filter shelf, marks card, show me(very popular
visualisations will show up here)-24 charts are popular
1}selection on data pane, tableau will activate certain set of charts- create bar
chart , text table
TABLEAU follows best practices in visualisation

when you go to show me ppanle bottom section ,min fields are important , how many
measures we need , how many dimensions we need
(example, profit vs regions)
column shelf will get aligned along x axis, row shelf will get aligned on y axis
you shoudld refresh this extract manually if you do it on tableau desktop, however
you upload it on on tableau server, you can set the frequency(monthly ,daily,
hourly with the help of scheduler )etc.
The following view is built with help of VIzql server
Rows : profit, category
Columns: region
Joins: Inner join , left join, right join , full outer join
cross database(conncetivity) join (coffee chain)-ms access and im joining with my
sample superstore(excel file)
orange bars, blue bars , tableau is smart enough

change the name of combine connection to some custom named name


create folders
group by folder
hierarchies parent child explaination (drilldown and rollup features)

sample superstore datset


global superstore dataset
coffe chain ms access dataset

You might also like