Tableau vs Python - Building a COVID tracker dashboard

Key Takeaways

so at this point i have come across so many posts comments on my blog post on different forums on my videos that you know tableau or excel or power bi these sort of tools turn out to be better for data analysis or creating dashboards of different kinds now i do not use excel or tableau you know for my regular data science work but i am actually working on creating some really solid job ready projects for data science aspirants and i decided that tableau or even you know or power bi or excel should be part of that curriculum and i started fiddling with tableau and i in order to be able to tell the difference or compare programming with tableau i thought of creating something that i've already done in the past so last year around the same time i created this cove dashboard using python jupyter notebook and voila so i decided to create the same dashboard replicated in tableau and what happened next to know more about it you should actually continue watching this video not only to find out what happened but to actually learn how to create a dashboard using tableau in just 15 minutes so the first step is to basically install tableau onto your machine so i hope you have already done that now i've already installed tableau public onto my machine so you can see this is the interface that you should be all seeing when you open up tableau you shouldn't be seeing this this is because i was already working on this notebook where i decided that i should do this video so this is that particular notebook and apart from that i think everything should be the same on your interface as well now the first step is to download the data so this is kobe 19 data repository which is maintained by the center for systems science and engineering csse at john hopkins university so they are maintaining this repository all of the stats and all of the numbers coming in from all over the world are being maintained in this repository and the data is uh you can find the updated data so i created this dashboard uh a year ago uh which was using jupyter notebook and python and i used a bunch of so use the same repository this is the same these are the same links so uh for our particular use case in this particular video would be to create a country specific i want to visualize the scovid situations in different parts of the world so i'm using this file which is giving us the data of of different countries from all over the world i'll simply be using this so let me quickly show you what this data actually looks like so this is the raw github file it's a csv file you can say cases country and it has got country region last updated latitude longitude of those countries those places confirmed cases deaths recovered active incident rate mortality rate so on and so forth so there are a bunch of things so this is actually serving our purpose or what we want to actually visualize this file gives us that data pretty simple we want to keep it simple and this is the dashboard uh that i made a year back which was using voila japanese book so even if you update it today it will give you the updated numbers because the data is all coming in from those files and i'll provide the link in the description uh to the repository to the video and to this dashboard as well you can uh check it out there so now comes the question uh how do we load this data into our interface so let's quickly go through a couple of things so you have this discover option on the right uh in this right pane where you can actually go through and learn a few uh you know tips and tricks and or how to work with tableau now whenever you start off a project the first thing that you do is load your data onto the software that you have so here uh there are a bunch of ways you can load your data so you can upload these excel file text file json so tableau provides support for uploading a bunch of different types of data and you can also connect to a server so let's say your data is you know available on an s3 server or it's on google cloud storage or you can say a simple google drive or in our case it's on it's a github raw file so it's on github servers so here it gives you this option for connecting with a web data server so here there is there are basically a number of web data connectors available for you to connect and you can actually write your own as well so you click on this web data connector you can see you can use a connector which are already available you can learn about what this is and you can also build your own connector so i googled as in how do i upload this csv file and to my surprise there was someone who asked the same question in the community so in the tablet community there was this uh guy or girl i think hey they asked this question how can i connect below csv and they also gave us the same link i think they're also using uh trying to do something with the kobe 19 uh building a dashboard or visualizing some finding out some insights from the data so and kisha rose so she has written this uh web data connected to connect uh csvs and we're going to use this she has provided this link to her heroku app and we are simply going to copy this link this is her web data connector and what you do is you simply put the link over here hit enter and it would load the interface and here you can provide the link to your csv file so let's quickly copy this link put it in here and click on get data so tableau would take a few seconds and load it so you can see i have the data here country region last update so this is the preview of the data that we have coming in from the csv file that i just showed you and this is the interface just after uh uploading or loading your data source so our data source is you can see on the left bottom so this data source is rightly set up so now what i do right after loading the data set is do some initial exploration and i try to comprehend like what sort of questions can be asked from the data set so here if i see i've got country or region which basically serves a purpose that i want to learn about the situation of kovid numbers number of cases now i also have the number of cases so i have got confirmed debts these are all number of cases that are coming up all around the world the number of active cases what is the incident rate what is the mortality rate where people are dying more and you know you have the latitude and longitude i can actually plot these numbers on a world map or on a globe or something like that and after that i have latest uh last updated this is all the same so won't make much sense over here to use this so yeah i think uh the kind of question that are coming to my mind is you know i can give a total number of cases let's say confirmed uh total number confirmed cases in the world total number of deaths total number of recovered that would be a brief overview of the total numbers that we have uh the some aggregate basically and then the next question could be what is the current state of the countries so that could be the world map i i can visualize that in one frame and tableau gives us interactive maps so that would be great as well and then which are the most affected countries in terms of the number of cases or the mortality rate so i can choose whatever figure or whatever sorry whatever feature or whatever variable i want to visualize and accordingly i can pick it up and plot it against the countries and i can also sort them so i can find out like which are the most affected ones so let's say the top 10 or top 20 or whatever threshold that i set in order to shortlist those countries that are most affected so these are the questions that i'm actually trying to find the answers to and once you have done some sort of initial exploration on your end you can define some different questions or also you can use some sort of some other file that's also available so now after this point i would click on this sheet one at the bottom here so once you click on sheet one the interface that you get over here so let me quickly tell you uh these are different there we get different sheets so you work on different just like in a spreadsheet or an excel sheet you get sheets here also you get different sheets and this is the main view you need to simply pick up a field and drop it here like this and give you uh it's currently giving us the sum so let's quickly delete it uh here you you can simply everything is simply drag and drop so you can pick up any variable any feature that you want you can draw a drag it and drop it over here we'll see how we can do that in a bit so here you drop the columns here you drop the rows that you want to visualize and these are a few marks so you can decide like i want to add something as a label or i want to add something in the tooltip or it should not appear on the visualization or the chart that i'm actually displaying it should be up it should appear in the tooltip so those are the kind of things and if you hover over any of the icons long enough it'll show you just like it's showing us color drag dimension or measures here or click for more updates so uh this is the data pane on the left so this data pane uh basically if you see you have blue uh these are the blue dimensions and the green ones are basically measures so dimensions are basically your categorical variables and measures are your quantitative variables so that's a clear distinction that we have that is uh done by tableau for us then other is other icons that you should know about is uh this is for sorting so this is for sorting ascending this is for sorting descending okay this is for highlight uh this is to show your mark labels so if you want to show something show the labels uh you can simply click on it and the charts that would appear uh they would be all labeled so we look at this in a bit so the this is basically what we have and also one more important thing is these are the number of charts that would be that you would be able to plot and tableau is very intelligent in that sense so that you know if you want to display let's say this world map so it tells you that for maps you have to try one geo spatial data so because you need to in order to plot a world map you need latitude or longitude so you need that uh you need zero or more dimensions for the it's it's going to be you know besides longitude you need to have latitude or some sort of a combination and then 001 measure so what are you actually trying to visualize on that world map so that would be your measure so if in the in our case it can be confirmed cases deaths or whatever so let's quickly get down to answering our questions that was all about the tableau interface you can learn more about the interface or using the links that i provided in the description or from the documentation that tableau offers so the first question is that i simply want to look at the total numbers so i'm looking at confirmed cases active cases and depth cases so all i simply need to do is this is my confirmed measure so i'll click on it i'll drag it and i will simply drop it in the columns so here you see we have million so the scale is in millions here and it's going uh let me quickly shut this off it's going close to like uh 120 millions and if you'll hover over it you see it gives us this tooltip which actually tells us what's the current number so the confirmed number of cases is close to 119 millions now i want to also visualize let's say the active cases so i'll pick it up and i'll add it in the columns so here you see these are vertically separated because i'm adding it into columns you can also do that so in in the rows as well so here you see these are like horizontally split so i want them in columns because it would take less space in my dashboard so again and then the next step is to put death cases as well so these are the three numbers that i have now you see i have to hover over it to actually find out what the current state is or uh what are the number of active cases confirmed cases death cases in the tool tip uh and there's nothing actually on the chart so let me quickly click on this show mark labels so i'll do that and you see it is showing me now this number has popped up so it's showing me the number of cases that are currently there for each of these types of cases so uh this is now looking better so you can also color code them so what you can do for that is uh simply go over here so you have these three uploaded in the marks you go to color or you can simply click on this one and then go to color uh let's say for confirm let's keep it to this one and for let's say active uh let's add another color to this maybe orange okay and then for our death cases uh these should be red so i have added some colors here you can add other uh formatting options as well so i've added some width to this as well so you can right click on it you can format and you can change a bunch of things about the rows the columns uh the entire sheet uh font or whatever tool table title uh whatever you want to add so that was all about the total number of cases now we can actually visualize the total number of confirmed cases total number of active cases and total number of debts uh you know all over the world you can next step is to actually rename this sheet so you can actually rename as well or you can you know enter that this is like the total number of cases and open up over here so you can also change the name over here so edit title you can say uh total number of cases uh press ok and you have total number of cases uh on the top now coming to the next sheet so we'll be working sheet by sheet and then at the end while building a dashboard you can just simply compile all of those sheets in just one dashboard so here is the icon for new worksheet if you'll hover over it it'll tell you that this is the new worksheet and if you'll hover over the second icon it is it is for new dashboard and similarly this one is for new story so here uh let's quickly click on sheet2 uh click and create a new worksheet now the next question that we wanted to you know visualize or answer was what is the situation of covered in all over the world so i want something our kind of like a visualization maybe a world map or something because we have dimensional geospatial dimensions uh data uh latitudes and longitude so we'll try and make use of that so here uh you see tableau has generated these latitude generated and longitude generated for us so all you simply need to do is pick up the longitude and drop it here in the columns so this is what we have right now and the second step is to pick up these latitude generated and drop it in rows so now if you see we have this very blank world map over here and i want you to click on show me and you see all of these are gray and except for this one uh element one chart and this second chart which is the text tables so you see the world map so this is the symbols map uh so it'll basically put on drops or circles uh on different parts of the countries so different parts of the world basically so here we have for the symbols map also you need one geo special dimension zero or more d dimensions and zero or two uh measures so let's quickly uh try what do we want to visualize over here so currently we only have numbers so we have latitude and longitude so let's provide names to these so i'll drop country and region i'll put it in the details or you can simply put it over here in marks and now as soon as i drop this country or region dimension in this marks you see i have got these dots from on this particular map so and clicking or basically hovering over any one of these circles it will tell you the country or the region that it belongs to and it's it is a basically a detail that we have added to this map now we don't simply want to look at the country names we want to visualize something on this world map so let's pick up let's say number of deaths so we could pick up debts and we drop it in terms of labels okay drop it over this label box so what it'll do is it'll add numbers so you click or you hover over it it'll tell you the country's region and it will also tell you the deaths which is 158 607 in india now for india you can see that this label is also available so since we dropped the death column on to the label box the numbers are actually visible if you drop something let's say i drop active on details so that is a detail added to your tooltip so now in the tooltip you see country region india active cases two hundred ten thousand and that's is the label which is also available which is also visible uh like that now this is how you can actually uh visualize these numbers and now if you see the other chart which is the maps simple maps this is also available so if you click on it this would create a chart like this and here also you can actually click on mark show mark labels and show those numbers onto your dashboard on those numbers or i am liking the symbol short move so let me do that so this is actually showing you so now if you see the shape or the size of those circles have changed a little bit as well so this is how we can actually create a symbols map to actually visualize what is the current scenario of each of those countries so we have displayed the depths on each of these places and simply you can add more details to your tool tip let's say i want to add mortality rate as well so now when you hover over something hover over a particular country or region it'll give you everything mortality rate deaths active confirmed kiss and so on and so forth so let's go down to the third question which is also let's quickly rename this one as well [Music] okay now click on the next sheet click create another sheet and in the next sheet now we want to visualize uh like the most affected countries in terms of confirmed cases in terms of debts in terms of mortality rate or recovered rate or you can pick up any variable that you want to visualize and this one is pretty simple so first of all you pick up the country or the region put it in columns and then the next step is let's say i want to visualize confirm cases so here you see we have got a very quick bar chart uh for us so this bar chart is pretty easy but we simply want to shortlist or visualize or you know understand the cases of the most affected countries so for that we need to sort the data so we'll click on sort descending so this actually gives us the sorted bar chart so here you see clicking over each or hovering over each of these bar will tell us that this is us which contains which has the most number of confirmed cases as of now and if we click on show mark labels we have got the numbers as well now we are not concerned with all the other countries so we simply want to visualize the most affected ones so for that in order to shortlist you simply uh you know select the numbers select the countries that use or the bars that you want to shortlist or screen and just hover over it and it will display this pop-up this tooltip which basically tells you that do you want to exclude these or do you want to simply keep only these so i click on keep only and this is an uncluttered bar chart that i've got for to visualize my most affected countries in terms of the number of confirmed cases now similarly i can do this for other variables as well so again i can change the colors of each so let's let me do this for another i created another sheet now put country region in here in columns then i want to add let's say mortality rate so these are the mortality rate i sort these i want to keep only these and then i want to add let's say or i want to simply add color so these this would be my mortality rate let me put it to red so this is my mortality rate uh i can name it and simply you can rename the sheet as well and rename the other sheet which is confirm click another sheet you can actually visualize the recovered cases as well [Music] so now we are done with all of our charts all of our visualizations uh so the final step is to create a dashboard now in order to create a dashboard you simply need to click on this new dashboard icon click on it and here you see you have got all of your sheets that you have worked on so these are all of my sheets over here and it has simply picked up these sheets and put it in the sheets pane this is my main view where i am required to drop all of my sheets in whichever way i want i can position them as per what uh you know how i want my dashboard to look so let's quickly do that one by one you pick up the totals and you put it let's say on the top then you pick up world map you put it over here you can actually change the shape or so let's put confirmed cases as well okay click on this one i want this on the top mortality rate put it over here recovered put it over here over here and that's put it over here so this is so now i simply need to position these so that takes a bit of your time [Music] so this is how i have positioned my dashboard you can drag and drop or you can put something else in here that's totally on you uh whatever you want to visualize you can accordingly position all of your elements it's simply about dragging and dropping so total number of cases on top right giving us uh the brief overview of in terms of numbers then we have world map uh you can simply you know these are all interactive elements click over any one of these and you'll get the numbers number of active confirm mortality rate death and then we have confirmed so you can see here you have the most affected country in terms of confirmed cases death cases recovered cases and mortality rate so that is how quickly you can create a dashboard out of tableau now the last step is to click command s or you know save this whole dashboard so for that you would be required to sign in so make sure you have an account if you do not just create one so add a workbook title so this is going to be my covered dashboard so save it and it'll be publicly available you'll be able to share it with your friends then they'll do some take some time to send all of your data onto the tableau public server takes a few seconds and this is the link that opens up so here is the dashboard that we created you can see a visualization author har shithyagi and similarly you can create yours and share this link with others and they can check out your dashboard all of the elements are interactive you can look at them feel free to share your dashboards uh and whatever you create in the comments down below i'll take a look and maybe announce a winner uh in the next video so so that's how easy it is to build a dashboard using tableau now after creating this dashboard and comparing it with the amount of effort that it took me to build the same dashboard using python and jupyter notebook last year i tried to score the two methodologies on different metrics on a scale of one to five where five is the best and one is the worst so the metrics that i chose were time consumption uh ui of the dashboard customizability and ease of creation and turns out that tableau seems to be a wise and time efficient choice over here at least for these use cases now this is my personal opinion and it may vary and depending upon your workflow your problem statement your use case these scores would change definitely so again if you like this video if you found this useful please do give it a thumbs up subscribe to the channel hit that bell icon and comment down below what sort of videos you would want to see on the channel or if you have something you know something to chip in around the same topic uh please please feel free to comment and share it with your friends your fellow analysts and i'll catch you guys next time