Tableau for Interactive Data Exploration | Visual Analytics with Tableau | Tableau | Edureka Live

Key Takeaways

hello everyone good afternoon welcome to this webinar where we will try to understand about Tableau okay what is Tableau and uh what is its purpose how does it help us so here the agenda of today's discussion is to understand what is data visualization why do we need Tableau how does it fit into the place what is Tableau how do we install it understanding the features user interface we'll also very quickly create a few charts and uh learn about a few features and functionalities of tblo okay and uh what is the path that we follow at urea for Tableau it is a program which consists of 12 modules in total we begin with data preparation using tblo prep Builder a tool used to prepare the data for analysis then we look at data connections using Tableau desktop then we look at basic visual analytics where you will learn how to create some basic charts and you will also learn certain Concepts like how to sort the data how to filter the data such things then we have calcul ations in Tableau post which we will be looking at Advanced visual analytics looking at the analytical capabilities of the tool like prediction forecasting clustering so on then you will be looking at level of detail Expressions which is also a part of calculations itself then there is geographical visualizations everything related to geographical Maps Advanced charts then dashboards and stories and finally one indust great project that we will work upon now it doesn't end there this is the 10th module then we have 11th module where we focus on Tableau online and 12th module has a small project that we will be working upon together in the class okay so this is uh how the course is structured now why do we need business intelligence forget business intelligence why why do we need intelligence and what is intelligence right see throughout our life since our childhood we are taught a lot of things by many people and we even you know go to educational institutions like schools colleges and learn a lot of things we gain a lot of knowledge so whatever knowledge we have gained through different media right through Eyes by watching things through our ears by listening to things smell touch whatever it might be when we apply this knowledge for the betterment right to apply whatever information we have gathered whatever data we have gathered whatever knowledge we have acquired when we apply it in such a way that it improves our lives and the lives of people living with us our our you know surrounding and our environment we say that it is an intelligent application of the knowledge isn't it so when you take all the technical skills that you have learned and when you apply it on a business then that is basically business intelligence taking everything and applying it to a business so that you boost the productivity um to gain you know monetary benefits to be able to reach the Target that is business intelligence and Tableau is one of the very popular tools that are there under the platform of business intelligence tblo and powerb both of them are data then why do we need data visualization because these tools primarily take huge amounts of data and they convert them into Graphics Graphics which are easy to understand and interpret why do we need this because picture speaks a thousand words I think as simple as that when you represent something in a graphical format people tend to remember it for a longer duration of time people can understand it better so on and so forth now look at this quartet this is called ancom quoted which was um which is basically four data sets in tabular format so X and Y coordinates are given and when you look at the basic summary statistics like summation of the data in column X and Y average of the data and column X and Y okay something called as standard deviation of the data and column X and Y you can see that the summary statistics are the same for all the data sources but they are all not the same different sets of data with same summary statistics because when these were presented in a graphical format that is when people first realized the power of data visualization because each data set had a different story to tell you can see data points going in a kind of an upward Direction This is a curvy linear pattern here it is perfectly linear with one outlier odd man out one outlier which is kind little bit throwing off Shifting the overall direction here all the x coordinates are 8 8 8 88 except for one outlier so instead of being a vertical line because of this one outlier it got thrown off like this each of these graphs have their own significance each of them is telling a different story about the data that we are looking at so just because the descriptive statistics or the summary statistics are the same we can't conclude the that the data is the same data is completely different here so that is when people realized the importance of data visualization okay scope of visual analytics it is pretty seamless it is basically for better financial performance better customer relationship that we can establish when we know the customers behavior and their purchasing pattern better decision making okay better sense of risk better key strategic initiatives that we can take up so data visualization is all about Gathering data from different data sources analyzing that data and presenting your analysis in a visual format presenting your analysis in a graphical format okay why do we need Tableau because it is one of the market leaders when it comes to creating interactive visuals it is an intuitive platform and it is very very easy to learn and very easy to create the charts using Tableau it is a software company which designed this tool called as Tableau and now it has been acquired by Salesforce quite quite some time back maybe two three years now Tableau is now a Salesforce company it was acquired by Salesforce a brief overview into the architecture of Tableau see um so input we take the data from different data sources let's call this input okay we're taking the data from different data sources and we process that data here process that information by creating different types of charts using that data and then after you have created the reports the output part output means what where you're sending all those reports to others for consumption so you can send it to individual users you can send it to a large group of Audience by uploading this those reports onto the server okay so we take data from different data sources we process that data and create charts and whatever charts are created can either be published onto the server this part here is Tableau server or they can be sent out individually to various users okay that is the purpose of tblo to take data create reports and share it with the stakeholders now from within tablet we can connect to multiple file formats and data from multiple servers not just simple files you can connect to data from multiple servers so suppose you have to connect to an Excel file okay then you just need to choose Microsoft Excel option and then select the Excel file immediately we are brought into this page in Tableau where we can see the name of the connection we can see the list of worksheets present in that particular connection and by dragging and dropping the sheets over here we can use the data present in those sheets okay now there are different data types that Tableau recognizes it recognizes Boolean data a field that might contain true or false values dates short date long date everything it recognizes date along with time stamp okay um string data alphabetical and Alpha numerical data decimals whole numbers and also geographical data after we are U you know connected to a data source when we come to the worksheet the basic workspace okay this area is called as the canvas the place where we see the charts take a shape is called as the canvas we have the menu bar then we have the tool bar this is called as the data pane where you can see the fields segregated into dimensions and into measures we'll talk about them then you have certain shelves like columns and rows there is something called as a marks card which can be used to work with the color of the visualization size of the data points label them okay customize the tool tip Etc so once you connect to the data source the fields from the connection are divided into dimensions and measures so what is a dimension and what is a measure Dimension is basically categorical data okay and measures are mostly numerical values on which you can perform any sort of mathematical operations okay and when you perform some sort of mathematical operations like addition subtraction multiplication division uh calculating the average calculating the median standard deviation variance whatever sort of mathematical operations you perform you will always get a meaningful result so mostly numerical values which can be used in mathematical operations go under measure Spin and rest of the data like who is the customer what did they purchase when did they purchase categorical data that all goes into the dimension Spain there is something called as show me panel which is available which in tableau's own terms is called you know the crown jewel of Tableau it's a very important tool it's like your best friend you know it it will help you decide which chart to create and even if you don't know how to create that chart just with a few Mouse clicks selecting the fields and choosing a graph here you can create the desired chart very easily we can create charts using the show me panel I I'll give you a demonstration of that so a variety of visualizations can be created using Tableau in a very very easy way bar graphs line graphs combination of bar and line graph so on and so forth let me let me take you to Tableau and show you first I will show you the connection and then we will see how to create some charts very quickly see if you would want to install tblo go to tableau.com which is the official website of Tableau and from there you can download a tool that is called as tblo public it is a free software anybody can download it and install it and you can use it for free forever if you go to this link of course you will be prompted to fill in a form okay you have to fill a form where you will have to give your first name last name and your uh personal email ID will work okay your personal email ID and once you've done that you will be able to download the tool and install it it's a very simple installation wizard you just click on accept the terms and conditions next next next and install okay now what I am showing you is Tableau desktop the professional Edition which which is the licensed version not the free one so here I will select Microsoft Excel and then simply choose the data source to which I would like to make a connection here okay I would like to use data coming from Microsoft Excel sample Superstore once I do that and click on open you will notice um how we are connected to super store which is an Excel file and we can see the list of worksheets present in it all I need to do is drag and drop the sheet here so once I drag and drop orders here you can notice that Tableau basically gives us a preview of the data on the data grid now we will proceed to the worksheet okay now we're going to Simply proceed to the worksheet and look at the canvas the data Paine Fields segregated into dimensions and measures so on and so forth so most of the things are in Tableau are drag and drop you just need to drag and drop okay I brought category and subcategory I I'm interested in looking at the profit job done look at that bar graph as simple as that so I created a bar graph which is used for comparison to compare the performance of members against each other now when you have time series data when you have data for over a period of time then you might want to create a line chart line chart is for time CDs data and you can notice how tblo automatically created a line chart okay suppose you don't know how to create it you can just choose the necessary fields and go to the show me panel and click on the chart that is recommended this is tableau's recommendation outlined in red color once I click on it chart is created it is that simple to create cre a chart now what if I want some dual access chart means what let's say for different subcategories I would like to check um the sales and I would also want to check the profit and I want to compare them then we can merge both of them into one graph like that okay so once I merge them with dual access concept once I merge them you can notice how everything has become circles now we can go to each marks card and this is called Mark type okay you can choose what Mark type you might want to use just go there and choose the mark type so you can have two different Mark types combined together like that this is a dual combination graph okay a quick demonstration of different charts that we can create here you can even change the color suppose I don't want orange I'll click in the color shelf and choose a different color Okay edit colors and choose a different color maybe I want this color for uh sales this color look at that and then profit I I'll just click on color edit the colors I want a red colored line for profit look at that okay we can edit the colors to match what is given so next geographical map and then area a graph with dual axis okay so if you have geographical data which is indicated with a globe icon against the field when you take Geographic fields to the detail shelf Tableau will automatically recognize them as geographical values and it will create a geographical map it is that simple again or if you're not sure what to do just select the fields and open the show me panel and select the chart whatever chart you might want you can create something like this or like this anything is fine so that is basically how you create a map let's see an area chart across different months okay across different months we're going to look at the sales and we're going to look at the profit so we have a sales line and we have a profit line now we can merge it with dual axis two lines you can even go ahead and change the mark type to area look at that okay we better synchronize the axis then we get a better Clarity look at that synchronized axis meaning it will make sure that the range of values on either axis are one and the the same we have the same range of values on both sides so you can see what is the profit and you can see what is the uh profit like sales is in light blue profit is in red see I had changed the color here right so the same color is getting carried on whenever we use profit and sales you might feel that it is very light you can just increase the opacity it becomes dark so this is basically an area chart this is called a heat map heat map is like a text table where you arrange everything into rows and columns and the information is conveyed like this through color without putting the actual values you're conveying the information let's say through color like the through the temperature of the map to the through the temperature of the map that kind of a thing so I'll just replicate the heat map and show you for various segments across different categories and different subcategories we are looking at the profit information so profit went into the color shell this is basically a heat map so wherever I see blue color or we can even change the colors let's go with red and green uh diverging so wherever you're seeing red or dark shad of red it means very bad loss is coming from there and wherever we seeing green okay it means very high profits are coming from there lighter shades of green meaning lesser profit so through the color intensity we are basically conveying the information here that is the purpose of a heat map you could use any color scale okay next is a tree map when we have hierarch data tree maps come in handy so suppose um there are different regions and across different regions I would like to check the sales okay this is a tree map this one I will color it by region okay so the size of the block is indicating the sales West Region with a very big block is making the highest sales then East then Central and the least sales is from South now each region contains multiple States so if I take state in there you can see how each block is divided into subblocks showing the performance of each state so west region I know that California in California the performance is very good from east region I know that from New York that the profit is very good we can even take state to label look at that further if I want to granular eyesee I could do that I'll bring City to label within California this whole block which corresponds to California we can see the performance of various cities Los Angeles and Francisco San Diego so on and so forth within Washington okay you can see the performance of different cities so on and so forth so this block corresponds to Arizona this block corresponds to Oregon this black block corresponds to Colorado so on and so forth within that you can see the performance of different cities so this kind of branching out of the data comes out beautifully when you create a tree map okay when you create a tree map it comes like that branching out of the data so bar graphs are used when you have to perform comparison line graphs when you have time series data and would like to check the trend dual axis when you're trying to compare two measures performance against each other when you have geographical data then that M okay again whatever dual axis it might be a regular one or area it is for comparison between two measures heat map is to visualize the variations across categories and tree map is when you have to show branching out of the hierarchical data okay so let's look at some functions of Tableau what can it do from Tableau you can perform joints and you can also perform Union suppose your data is coming from multiple tables which need to be combined all the columns from both the tables should come together resulting in a big table with all the columns put together then we perform a join for instance over here I have orders which has 21 fields and I have people which I can just drop here and Tableau will automatically join it we have two columns in people which are combined with 21 Columns of orders so people came and got appended here okay suppose I want to bring returns also then you see here returns people orders in total 25 columns the number of rows will depend on the type of join that you apply you can even change the join operator very interesting feature combining data from multiple columns and joining them to perform some sort of analysis on the data okay not just this suppose you have a scario where the data is simply spread across multiple files okay it is the same data same structure it is just stored across multiple sheets or multiple tables and if you would want to upend the data like 2011 transactions are there and I simply need to append 2012 there Union you can perform a union the number of rows will increase up in 2013 look at that so we can perform join combine ing The Columns getting and the table gets extended horizontally or we can perform Union where data from two tables are appended together like this it gets extended vertically depending on the scenario depending on the business requirement you can choose what to do whether joining the data would be the right thing or whether performing a union on that data would be the correct thing to do apart from joints and unions there are other things like establishing relation a ships between the data sources something called as data blending which will work at aggregate level after you have aggregated the data you can perform data blending so these are all capabilities which Tableau has with respect to manipulating the data because data could be coming from multiple tables from multiple data connections you may have to combine it so for combining we have various options be it joints or unions or relationships or data blending cross database connectivity it has a lot of capabilities okay it's pretty extensive when it comes to working with data next we will look at some features okay which is like sorting can we sort the data we can do that so for example I have category then and I will take subcategory also and let's say I'm looking at the sales information okay and I will color this by category now look at the sequence you have data for different subcategories under Furniture members under office supplies members under technology but they are being arranged here in alphabetical order isn't it they are arranged in alphabetical order what if you might want to rearrange them in a different sequence so you just have to click on this sort button on the toolbar data gets sorted in descending order if you click on this one it is sorted in ascending order the opposite so sorting the data is as simple as one click on the toolbar nothing much just one click on the toolbar then there is one concept called as creating a set which is again a very interesting feature because set will help you to Define conditions and members which meet the condition that you define will be kept in the set and members which do not meet the conditions or the criteria that you define they will be taken out of the set so there's a concept of keeping the data in the set and the concept of taking the data out of the set in and out okay so for example here for various months across different ship modes we looking at I think profit here okay yeah ascending and descending is based on data you can see the longest bar which is having the which is making the highest sales at the top of the list isn't it so this is based on the sales values that these members are getting rearranged it is by data so the chart is like this for set the chart is we have year and we have month we have year 2012 13 14 and we have months listed out first class same day second class this is the shipment mode okay and so we'll be looking at the profit here and it's a DOT Circle plot so you can see circles indicating the profit okay something like this now we would have offered some sort of a discount isn't it so discount I will take to the U detail shelf or tool tip just show you all what is the average discount being offered so over here uh let me even format it to represent it as a percentage okay without any decimals all right now look at this so on an average what is the discount offered average discount is what I'm showing but our requirement is we are going to create a set where the discount is greater than 10% okay so I'm going to create a set I want the members where the discount offered is greater than 10% okay so I'll give 0.1 now of course I could have renamed it now if I take it into the color shelf it is getting segregated you see where we have given an average discount of less than 10% and where we have given more than 10% so wherever the discount is less than 10% it is in the set okay this is at um um row level okay this is at row level so what did I say just a moment please discount should be less than 10% no greater than 10% sorry sorry greater than 10% so wherever we have given a discount of greater than 10% so you can see discount greater than 10% and then less than 10% you can see a gray colored circle that has come out okay so wherever it is available you can see the data point wherever we have not offered less than 10% it is not there greater than 10% is in the set okay discount not greater than 10% is out of the set so that that is what the chart is indicating right so we can create sets based on any condition and members which meet the conditions that we specify will be in the set members which do not meet the condition specified will be taken out of the set so set is very Dynamic it it is based on conditions it keeps checking the condition at any given point of time members which meet the condition will be in and members which do not meet the conditions will be out they will be taken out okay then there is something called as forecasting that we can do so for example now what is for casting for casting is where you like look at the old information you look at historical data data that has been gathered for over a period of time and using that data when you um decide or when you try to estimate what the future values will be like so when we look at the historical data and try to estimate what the future value vales are going to be like that is called as forecasting and Tableau has an inbuilt capability to perform forecasting on the data for example let's say I'm looking at the sales across different months and I would like to forecast for the next few months all I need to do is go to the analytics pan and drag the forecast indicator drag this forecast into the chart once I leave it there it will perform the forecasting and it will show me the estimated output you can see actual data in dark blue and estimate in light blue we can even further go ahead and change the forecast options decide what should be the forecast length meaning how far into the future are we trying to go okay and uh certain things like whether or not we might want to ignore some data points how we would want the data to be aggregated okay what kind of forecast model that we might want to uh use or apply it is exponential smoothing that it uses okay something called as prediction interval that we can set So based on these settings accordingly the forecast will get updated now after you have performed the forecasting and you have the estimated values coming up in light blue what if you might want to share these numbers with your client you can do that just right click on the sheet and there's an option called as duplicate as cross tab so the data that we obtained in this graph and what whatever forecast result Tableau generated it gets saved as a separate worksheet you look at that so estimated values the lower prediction interval upper prediction interval and what is our estimate so we saying this is what we estimate the sales will be like but it can fluctuate anywhere in this range it could be from 21 to 59,000 most likely to be 40 so we telling what is the lower bound what is the upper bound and what is the value that we expect it to be so it's going to fall between the lower and the upper B bound most likely to be this number okay so that is forecasting most of the things as you might um be noticing already they are drag and drop all the features are there you just need to know from where to access that feature what to drag and where to drop it drag and drop drag and drop finish and the charts start taking a shape and and features also the inbuilt uh Advanced capabilities also can be utilized by simple dragon R like for forecasting all I did was I dragged forecast and Dred it there okay there is one more uh thing that we're going to see how to create a text table and highlight the data like that so if you look at this it's showing the discount and sales for various subcategories and what's the color indicating profit how do I know because I'm actually looking here I'm looking at the color shelf and the color icon and the field corresponding to it so this graph is indicating sales and discount but it is colored by profit indicating sales and discount so how do we create such a chart this is basically a highlight table okay so for various subcategories we are looking at at sales and we are looking at Discount sales and then discount now we would like to color it by profit so profit to the color shell on small changes Mark type should be square look at that now if you're not happy with this color you can always change the color let's say I also want green and gold okay so like this we can create it they have reversed the scale like this so negatives are going green and positives are in Gold so I reversed it accordingly okay there is another feature called as highlighting very interesting feature we can highlight the relevant data in the graph for example for various subcategories let us say we are looking at the profit and also sales no not for uh sorry this is by month okay we looking at profit and sales across different months across different months now this is granular ised as you can see this is granular ised by subcategory so we have a lot of lines we have so many different lines indicating the sales and profit it's absolutely cluttered and it's very difficult to figure out which line corresponds to which member here we can turn on a feature called as highlighter Once you turn on highlighter what happens is you can just hover the mouse pointer and you can see the lines relevant to the member that we uh you know go towards relevant to that member so if I just go towards chairs you can see data corresponding to chairs is highlighted if I go to copers you can see the lines corresponding to copers or envelops or fastness relevant data gets highlighted in the graph and the rest of the data is dimmed out this is highlight feature so when you have a chart with too many data points and you feel it might look cluttered or people may not be able to properly understand what we are trying to convey we can simply do this okay after creating charts you can notice that we can have only one chart per sheet in Tableau there is one restriction you can have only one graph in each sheet if you have a bar graph here you can't have a pie chart or a geographical map something else over here not possible but we bring all of them together onto a dashboard dashboard okay and on dashboard building an active report is as simple as dragging and dropping things there you just drag and drop the graphs whichever graphs should be involved or should be kept over there just drag and drop them finish and not just that we can get there is something called as device designers it will help us get a preview of our dashboard on different devices so when I click here see default I can see what I can can get an idea of what this dashboard will look like on a tablet this is what it will look like on a tablet what will it look like on a phone you see it is automatically customized in such a way that the graphs are arranged one below the other so we can get a preview of our dashboard on different devices we can see what it will look like on different devices and make sure that it is not getting cluttered or it's not getting you know charts are not getting get overlapped information is being presented in a neat format so that everybody can understand what we trying to conve so device designer is a very useful feature it helps you get a preview of your dashboard on different devices it's basically an interactive report okay um so these are a few capabilities of Tableau in a nutshell like we can connect to different data sources we can perform joins we can perform unions we can create a variety of charts and ultimately we can put them together on a report so that is the beauty of Tableau and it's pretty easy a very simple tool where most of the things are drag and raw all right thank you all so much uh I hope it was informative all the best thank you bye-bye e e for