Learning Objective:
- Data Analytics and Predictive Analytics
- Convert Data into Knowledge
- Cleaning, exploring, visualising data
- Tools and versions
Data/Information/Knowledge
Data - In various forms. Visual, written, audio, video. Data is collection of raw facts and figures. Eg. 100% humidity in Chicago
Knowledge - Domain or derived (raining or going to rain. 24 hours after it rains in Chicago, it will rain in Delhi. ) Profit can be derived only from Knowledge - it is related to dollar, not data or information. - If it rains, sell raincoat and umbrella and make money out of it.
Information - Processed Data is information. (It will rain in Delhi, so necessary arrangement)
Process - very cheap, - We need Domain knowledge and Tools.
Tools - Power BI , Tableu are tools.
Business Intelligence give Data Perspective
Predective Analysis - What if.
What is BI or Business Intelligence?
- Set of methods, processes, architecture, application and technology
- that gathers and transform raw data into meaningful and useful information
- Used to enable more effective strategic, tactical and operational insights and decision making
Three steps, three layer and three uses.
It can be done by tools, processes and visualisation.
- Power BI is capable of processing the Data also. Extract-Transform-Load (ETL) capability and make report
- In Tablue ETL capability is only 30% but Visualisation is great. - More useful for presentation and story telling.
Why use Business Intelligence?
- Accelerating and improving decision making
- Optimizing internal business processes
- Getting to know about business past, present and future
- Increasing operational efficiency
- Giving quicker responses to business related queries
- No guess work
- Meeting/Exceeding Manufacturing and Inventory goals
- Gaining Valuable insights into customers behaviour.
No more customer satisfaction but customer delight. Netflix - recommending clients what to see. 70% users follow their recommendation.
Why not excel?
- Excel is a tool for data gathering
- Cant add data from different data source in excel and so cant optimise
- Increase operational efficiency
Process:
- Access the data - using SQL , CSV, Excel, Web, Folders - Part 1
- Clean up the data - Part 2
- Mash Up the data - Part 2
- Explore the data - Part 3
- Visualize the data - Part 3
- Share the data - Part 4 - Like whats app, others too need to have for right usage.
- Refresh the data - Part 5
How companies use BI?
Machine learning and AI can be imbedded into it.
Power BI
AR VR - Augmented Reality and Visual Reality
Tableu and sales force are now one
Power BI family
BI is a subject and Power BI is a tool.
Power BI:
We need to download the software. Link for Download: https://aka.ms/pbiSingleInstaller (kindly copy paste the URL in the browser)
This will be saved on your desk top, when you click this, you will find the green screen as below or an yellow screen, This has all tutorial and instructions.
Top - Tools
RHS - Visualisation
LHS - Canvas/Data/Relationship
Bottom - Page number
1. How to access different kind and format of data.
Three ways to get data:
- Home menue data tab - get data
- from the screen
Go to home menu - select CSV file, connect -
- three drop down - first two leave as it is - third,
- You can load data or transform data - Load when data is clean, transform when something need to be changed (power BI is not importing the data, it is only connecting)
- Way of representing data in power BI - on the right hand side - Table - Numeric, timeline and blank.
- On the LHS we have Report/Data/Model - You can select what you want - data (Is full data not review of data), base data is not changed.
Data is connecting - extracting and representing here. All text is left aligned and all numbers are right aligned.
RHS- Report/Query - Transform Tool bar - (Power query Editor) - Its ETL tool - Separate tool of MS incorporated here.
Its two different sheet now.
In the beginning of each cell, there is a set of three numbers and three alphabets,
Method of Undo is not cntrl Z - we have to cross, what we had done.
Load and apply.
80% of time go in cleaning the data.
Close and Apply on top LHS.
You will come out of ETL tab, back to one sheet.
Start working on what need to be done. - Here check Balance and Gender.
Note:
- there is no way to zoom up/down the canvas area; you can drag your graph
- Ctrl Z will reverse the last step here.
You can follow instruction on top when doing ETL
When you insert excel with multiple sheets - a navigator will open.
Transform - home - use first row as column.
To create page, click on + below near page 1, you will get page 2
Import a folder - How and why import a folder?
How - combine and load
Why - When you need to add, you will just need to insert the new file in the folder in the desktop and refresh powerBI.
Visualisation:
Ask questions before plotting anything, with basic concept of visualisation, give better plot.
Ask 5 Why's.
There are four basic presentation types:
- Comparisons - between countries or elements.
- Composition- How these two data points are contributing or helping each other - Pie chart - ratio of population of various countries
- Distribution - Statistical - Sales and Profit.
- Relationship - Growth of Population and growth of GDP
- How many variable do you want to show in a singe chart - Number of columns
- How many data points will you display for each variables - Number of rows
- Will you display values over a period or among items or groups - time factor - change in population, GDP wrt time.
If you try to accomodate all in one, it will not help.
Item, Variable and Category are important
If you want to compare between Sales, COGS and Unit Sold by country - which chart till be right?
Among item -) One variable per item -) Few Category -) Few Item so we select = Column Chart.
Category is country and Item is COGS, Sales and Unit sold.
Export to excel- table from three dots, and graph format - from file - to pdf
Numbers plotted will automatically come to Y axis.
Pie chart will be good only upto 8 slices. If it is more than that go for tree map. You can do two level of categorisation there. - this is to see composition, so there will be no X and Y axis.
Slicer is embedder - filter in the chart. Alternatively you can use dropdown. slicer has an option to have All - from format.
BI is different from Data Analytics - but a limited ability for analysis is available in power BI , which can help in some charts like line charts.
For Predective analysis, you need to use Assure. - you can write a code of R and Python.
No comments:
Post a Comment