Transcripts
1. Introduction: Hello everyone. My name is Tyler Pernes and I'm a data analyst and data engineer. Today, I'm going to talk to you about no surprise, data. Now usually when people speak and mention data in everyday conversation, they're talking about a specific finding, like 80 percent of people have [inaudible]. Then they mentioned why that matters. But maybe it's a specific opinion they have. Maybe there's an action they would like to see based off that finding. But what's not shown here is all the steps involved to get to that conclusion and that's what I like to go over in this course. I like to go over both sides of what I call a data process flow. The first side is going to be the science side. This is very focused on getting the data in the first place, structuring it, and making sure it's stored in a logical way. It's more the engineering side of data. Now once you have the science side for that you can focus on the other side, which is the art side of data. This level needs creativity a lot more. This is more about building visuals to ultimately gain value out of that data, which is really the whole point of data in the first place. In this course, we'll go over each step in this process flow. We'll use those findings by working on a project together that's focused on gaining clarity in your own life. Hopefully by the end of this, not only way will you have a good understanding of data in a very high level, you also have gain some good experience so that you can just try it with your own projects.
2. What is data, and who really cares?: Hey guys, thanks for tuning in for this course. As mentioned in the trailer, this course is going to get you started with data analytics, whether that's for your everyday life or for your business. In this video, I'd like to talk about a few things. First, what exactly is data? What's the whole point of using data? Then I'll go over each step in the process flow is on a high level, as well as the remaining periods are going to look like. During this entire course, I do suggest if you have any questions to reach out in the comments because I want this to be very interactive. So let's get started. First thing to talk about. What exactly is data? Data in its purest form, has a very simple definition. Information. This means the color of your shirt. This means any voice that comes out of your mouth. This means leaves behind you, every blade of grass that I'm sitting on is information and therefore data. The point is, data is everywhere. But the trick is to be able to filter out the data you don't care about, so that you can focus on the pieces of information that's relevant to you. But how do we know what data to filter out? Well, you need to make sure you always have an overarching question. This is ultimately what you're trying to answer by using the data. Let's say you're at home and you having a conversation with one of your good friends, let's call him Johnny, and halfway through the conversation, you realize something, you haven't spoken for 30-40 minutes now, and Johnny is continuing to talk about things that you honestly do not care about. That gets you thinking, "does Johnny ever talk about important topics?" This is your overarching question, and you want to infer this. So what you decide to do is in your next conversation, you take your phone, and you put on the record button, which means you record every word, both that you and Johnny and talk about this is you extracting the data from the source. The source in this case is going to be Johnny, and your extraction method is the recorder. You do this for a few weeks, maybe even a couple months, so that you have a solid amount of data that you can reference moving forward. Once you feel comfortable with the amount of data you have, you figure out it's time to actually put it to good use. To do this, you start listening to all the recordings you have, you start realizing something. Those are a lot of extra fluff and there is a lot of words in here that it's just not useful to answer your question. So what you decide to do is instead of typing down every single word on the recording, you pick out just the topics that he's talking about, and then you put each of these on, say, a sticky note so that you can have it stored for future use. This is you transforming the data in this case from the recording, in conversation form into one or topics that's stored on a sticky note. Now that you have all the topics, in sticky note form, and figured it's time to actually leverage these topics and use them to answer your question. The thing is you have about 200 different sticky notes here. It's really hard to organize it, so what you do is you open up a spreadsheet, let's say Microsoft Excel, and you start typing down each of the topics. This is your process of loading and storing the information. So now that you have a spreadsheet set up, you want to analyze, you want to figure out how to use it to answer your question. You open up the spreadsheet and you're ready to get started, but you get a low over room still, even though this is a more structured way of showing the data, it's still a wall of text. It's still not particularly intuitive what to do with it. So you take some time to look into it and you explore the data to see what it looks like, then you decide to make visual so it's a little bit easier to understand what exactly is happening. You built a few visuals and now you're ready to dig deep and figure out, "Does Johnny action talk about useful things?" You see a chart that shows his top topics or most used topics, and you see the top three are; celebrities who annoy you, types of hot sauce, and puppies. You also realize these three topics account for the vast majority of all the conversations. In fact, they account for 90 percent of our conversations. This is you in analyzing the data to gather specific findings. Now that you have these findings there, what does that mean? Well, besides thinking about it, of the top three topics, only one of them is particularly important, and that's puppies. The other two are really closer to that "who cares" side of the spectrum. So you categorize each recording. You do this for all the topics and you realize something. Seventy two percent of all of Johnny's conversations are based on topics that are not important. The other 28 percent, of that 28 percent, 99 percent of it is related to puppies. What this means then is that Johnny does speak about important topics on occasion, but it's almost are due to one topic of puppies. That is an example of you telling a story to help gain a conclusion to answer your overarching question. So quick overview of what we just did. We figured out what an overarching question is, we extracted the data from a source, we transformed the data into sticky note form, we loaded and stored the data into a spreadsheet, and then we explored and visualized the data so that we can analyze it to gain findings and board the story to ultimately draw a conclusion. We'll be going over all these steps in detail in the next few videos, and then we'll apply them through a project that I think is going to be particularly relevant to you guys, is actually going to be a project focus on helping you guys keep track of something very important, at least I would assume would be very important to you. I'm a huge believer that data should be used more often in our everyday lives because the whole point of data is to help us inform decisions, and that's the point. I don't see why we shouldn't use data in our everyday life more often, I actually think the main roadblock is that people just don't know how to get started, despite the fact that it's actually pretty simple, completely free, and not too much time investment.
3. Process Flow: Science Side: Hey everyone. In this video, we're going to get started with the data fundamentals from the start of the data process flow, which is the science side. The whole point of the science side is to prepare the data in a structured manner so that we can ultimately gain value from it by using science side, think of it as this preparation stage. But what do I mean when I say structure? Generally you want your data to be set up in a table form, which consists of the columns and rows. Columns are found at the top of the table and data find the data to answer the question, what does this data represent? For example, let's say you are shown a series of numbers on a spreadsheet without any column headers. We have no idea what this data represents until we see what the header shows. So if one of the columns says revenue, then it's very clear what these numbers represent. Rows on the other hand store the information in the table. They actually act as the data itself. If we have no rows and just columns, then we have no data, we just have definitions of data that could be in there in feature. Since you can always gain more information about a subject, the number of rows can and should fluctuate, both adding and taking away rows. However, definitions of data really shouldn't change. So because of this, the number and order of columns should stay the same. All right, so that's how a table is set up and this is what our goal is. We want to get to a point where we have a structure of consistent table that makes sense. So how did we actually get there? Well, let's take a look at a framework, that always comes from one or more sources, and this can be through e-mail or this can be through an application, this can be directly from someone's word of mouth. The point is data can come from a variety of sources. This also means that data can come in a variety of forms. Sometimes it could be table form. Sometimes we might just have a line of text that represent data. Sometimes you might have an actual voice memo or a coding that you have to use. Regardless of where the form is we first need to extract the data, and that's a very important step one. Once you have the data, we need to start looking at the structure and we need to answer the question, is this data setup in a logical way? Consider these two tables. The table on the right is our goal table. This is how we want our data to look like. The table on the left is our extracted table. Notice the difference between the two, in the goal table we have three columns, country month, and population. The same data is stored in the extract its file, but we have one column for each month, and populations are stored in each of those columns. Think about how this extracted data can fit within the gold data set. Can they fit together like a jigsaw puzzle? If not, we need a way to transform the extracted data so that it can better fit with our goal datasets. In terms of loading the data, this is just taking the data that you just transformed and putting it into the table you've created. If this is the first time you've had this data set, you'll likely need to first create a table so that you can then load the data. In terms of status storage, there's two different types of storage systems. First, data spreadsheet. This is more of a small-scale solution. It's very simplest setup. The benefit of using a spreadsheet, it's very quick. The issue with spreadsheets, you don't want to use it if you have too much data. Couple of examples are going to be Google Sheets in Microsoft Excel. The other type of data source system is a database. Now this is more of a large-scale solution. Takes a little bit more time to set up and requires more technical skills. It's a whole lot more robust and you want to use it when you have a lot of data. But because it takes a decent amount of setup and it takes a decent amount of expertise, we're just going to focus on using spreadsheets. That's it guys. That's the fourth steps on the science side of the data process flow. Now, one thing I do want to mention is how these steps can be extremely simple or extremely complicated. It depends on the number of sources, the amount of data, the frequency of data, and how complicated the data is in the first place. For our purposes, when we go over the project, we're going to have a very basic setup that you guys don't need any technical background for and can get started with additional projects using the same process. In the next video, we'll learn how to leverage this prepared data set by going over the side of data.
4. Process Flow: Art Side: Alright guys, at this point we have figured out how to prepare a structured data set. Now let's talk about how to actually leverage this data to reach our overall goal by going over the art side of the data process flow. To bring back an earlier topic, our goal is to answer our overarching question. Before we jump into this, I'd like to first talk about a specific type at tool, and that's a Business Intelligence Tool or BI tool for short. BI tools connect to a data set and allows us to build different types of visuals, which makes it a lot easier to understand what a data is showing versus say, looking at table itself. We'll be using one tool in particular during this course, and that's Tableau. The main reason I'll be using Tableau is simply because that's where I have experience with, but given that there are a ton of different tools in the space. So feel free to use whatever tool you are most comfortable with. I will say that Tableau has a paid version, but don't worry, we'll be using a free public version for our project. Let's look at the outside of the data process flow. Similar to the science side. There's four steps; explore, visualize, analyze, and story-tell. Before we even get started with the exploration stage, you first need to connect data to your BI Tool. There's usually an intuitive user interface to setup this connection. Once the data is connected, we can start investigating the data through the exploration stage. This step is all about testing out different ways to view the data. Simply drag and drop columns in your BI Tool and see what visuals look good. When looking at a visual, think about these questions. Does the visual provide useful insights? Is anything in this visual out of the ordinary? Does this visual relate to our overarching question? Not every visual you test is going to be relevant. So expect to throw a few of them out. After this stage, you should have a few relevant visuals to use for the remaining steps. The next step is to visualize the data by building a dashboard. A dashboard is a collection of visuals that provide a vast amount of information at once. It oftentimes comes with interactivity, which gives you the ability to filter out specific data or draw down into specific sections. Since a dashboard is natively connected to your data set. As more data flows into the data set, more data will automatically show up in the dashboard. Once you have a dashboard setup is time to actually analyze the data. This stage focuses on turning the dashboard you built into meaningful findings. What are the key takeaways from this dashboard? How does this dash would help support or overwriting question? Think about a few different ideas and don't worry about putting them together for now, just put the ideas down on some paper or some document so that you can let them fit moving forward, do note that it's better than having too many findings versus too few. If you're not sure if findings is going to be particularly relevant, don't look at them anyway. Once you have this list of findings, it's time to put everything together and build a story that'll answer our overarching question. Building a story in this case is simply answering, how would our findings relate to our question? How you show this story is completely up to you and typically depends on your audience. If your audience is going to be a customer or client. Is it going to be a specific class that may be you're teaching? Or you're just presenting something to a friend. Or maybe you even just showing something to yourself so that you can figure out what you need to do next. There is a lot of different ways you can show in building story, such as a PowerPoint, a presentation or just writing things down on paper. But again, think about your audience. Think about what the best approach is, and that can really guide how you actually present the story. That's it guys, that's all we have for The Art Side Of Data. Those are the four steps and at this point, we've gone over all eight steps of the data process flow, both the science in the art side. Now that we know every step, we can use these learnings to work together on a project, which is going to be what the next three videos we're going to be focused on.
5. Project Overview: Haptracker: Quick recap on where we are. We will be going over the data process flow step-by-step. Both the science side and the art side. Next up is the fun part. Let's apply our knowledge to a real life situation. For this project, we will be analyzing something very simple. Happiness. Specifically, we'll set up a project to track our happiness at different points in a day and build a dashboard to help us analyze it. For the purpose of this course, let's call this the happiness tracker, or Haptrack for short. The first step with any project is planning. To keep it simple, let's just focus on answering these four questions. Number one, what is our overarching question. For us we're just looking to increase our happiness. So the overarching question is going to be simply, how can we increase our overall happiness? Number two, what are the sources of data? Well, the source needs to understand how happy you are at a given moment. So there's really only one option here, and that's you, you are going to be the source of data. Number three, what tools are we going to use, and what processes will we have in place to leverage these tools. For the extraction phase, I suggest having an application that's very quick and simple to use so that you can easily jot down where your happiness rating is at any point of the day. Let's use drafts. This is a very quick tool, very simple. There's no low time and it is basic. We really don't need to get too crazy with the extraction tool here. Now, synthesis is also happening multiple times of the day. It may be very easy for us to just forget to input in what our happiness is. So I suggest putting a couple daily remind setup so we can easily get reminded, it's time to actually jot this down. For the transform, load and store steps, we need to have some type of spreadsheet. I suggest using Google Sheets. Now, what I like about Google Sheets versus Excel is it's connected online, it is collected to a server, which means it can natively connect the Tableau without you having to do any manual intervention. If you do store Microsoft Excel, you are going to have it stored locally, which means it's not going to automatically update when you do connect to the BI tool. That's it for the science. For the art side, everything's going to be using tableau. It's really not too much we have to go over here. I will say the exploration stage is a planning phase anyway. So we don't have to go into much detail. Now the last question is, who is the audience? This is simply just going to guide what your story-telling approach is going to be. So it's good to know this beforehand. The audience in this case is going to be yourself. Because of this, you don't have to be too crazy with your storytelling method. You don't have to have a presentation or a PowerPoint or anything fancy, as long as you have an idea of how your findings relate to you over a compression, then you should be good. That's it for the planning phase. In the next video, we'll put it to action.
6. Haptracker, Part I: Alright, let's start building on our project. In this video, we're going to build out each of the sign steps, and the next video is going to be focused on the other steps. Now, step one, before you even started with the extraction step, we first have to download our extraction tool, which is going to be the drafts app. Open up your mobile phone, go to the app store and download drafts. Once you have drafts downloaded, let's go into it. You will see a blank screen, simply tap on it and you can start typing in whatever you know, maybe, for us, I suggest us putting in a number so that you can very easily understand why your [inaudible] was at that given time. Now if you click on top left here, you'll see other notes as well as when was last accessed, both date and time. If you click on a specific note and go to the top, you'll see the information about it, including when and where it was created. Just some extra information that you don't have to type down. Last thing to know about, when you go into the notes section, you can swipe left and right to either trash or archive it. That's really all you need to know about drafts. Other thing to note is that if you do just type in a number and then lock your phone, it's going to automatically save it. You don't have to worry about having time to save button, but an autosave feature. Since this is going to happen so often during the day, let's set up a couple of lines just so that we can get a reminder for it. Let's open up a remind and you can see here, I just set up daily lines in the morning, and in the afternoon, and in the evening, you're more than welcome to choose whatever times you want to track, how often you want to track and so forth. I just use this because it fits well with my schedule. That's it for the extraction step. Next, let's focus on the three steps which is going to require us to use Google Sheets. Let's look into that. To get into Google Sheets open your browser and type in google.com/sheets. Now you use Google Sheets, you're going need a Google account, so if you don't have on make one, it's completely free. On the Google Sheets screen, you're going see on the bottom are going to be some of your existing sheets, but just click on a new blank sheet and it'll take you to the untitled spreadsheet. This is a spreadsheet, spreadsheets like to move around and maneuver to different cells or select multiple cells, and allows you to input data, whether that's numbers or text. For our purposes, we're just going to use a spreadsheet to build one table with three columns. Day. We're going to have 'time of day' and we're going to have 'happiness fading.' Now you know drafts and you're going to start typing in the information that's found in each of your draft notes. Let's say we have one note on 8/27/2017 for the morning that had a happiness rating of six. You input each one in accordingly, do the same thing for both the afternoon and for the evening time of days. Now I do suggest this is takes a little bit of investment to open up the spreadsheet, maybe do this once a week not to worry about doing it everyday. But one thing I do want to know is you have the ability to copy and paste by the right-clicking and hitting "Copy" or just hit "Control C," and then right-click to hit "Paste" or hit "Control V." This makes it a little bit easier to use when you already created, so you don't have to type everything from scratch. Last thing I want to talk about in the bottom here, they are the ability to add more tabs. This can be useful if you want to have two different datasets. For example, Johnny's data and my data, what's higher is that. Then you can change the title of the spreadsheet at the top. Let's call this "Haptracker for Tyler and Johnny." Now it's going to take some time for you to get a robust data set, since you're only going to be adding three rows at most everyday. What I'll do is I'm going to share my dataset so that you can get started with the arts side. You can see the dataset in the description. When you open up this dataset, you'll see data all the way back from the start of 2017. Click on this square at the top left, right-click and hit "Copy" and then open up your own or just hit "File," "New." Go into the first cell, and then right-click and hit "Paste." That's all you really have to do here to get the data. Again, this is going let you set up a dashboard with my data and then what you can do is, you can just replace my data with your data so that nothing really hasn't changed. Now that since you already built everything, everything's just going to flow in naturally with the new data you have. Alright, that's it for setting up the sign size for this project. In general, there was a lot of different ways you can set up this process. So as you have more projects moving forward, use whatever approach you feel most comfortable with. I would say there's a lot of ways you can automate this, but we're not going to go over this yet. We'll go over this maybe in a future video. Alright, let's keep going and then look at the art side for this project.
7. Haptracker, Part II: In terms of setting up the art side, everything is going to be focused on our business intelligence tool, which is going to be tableau. First step is download Tableau Public. Open up your browser, type in public.tableau.com and viruses in your email, just enter email here, click download the app and that should start downloading the In Stock. Here's tableau, you really care about the left side when he opened it up. This is the connection screen. Scroll down until you see Google sheets, which is stored in memory for me, you might have to click on the "More section" and you can see everything that Tableau connects, its quite a bit. You can also just search if you are overwhelmed. Just click on Google Sheets and it's going to pop up a login window. Couldn't open your sheet information. Once you're logged in hit "Allow" and then it's going to show all your dual sheets that you connect to. Now, I'm going to connect to this one. Hit "Connect" and it might take a second to connect. Here, now you see the actual connection user interface screen. Now you really don't have to worry about too much on here. It should automatically populate with your data. There are multiple sheet option if you have data in more than one tab and you can just drag and drop into this interface. Anyway, let's move on to the actual worksheet. Okay, so this is what the user interface looks like for Tableau school over on a high level. On the top here you see the data source itself and if you right click, you can see information about the data. You can view the data. For example, if you want to see how it looks, which should be the same as your Google Sheet. This section, are your dimensions and measures, which is just another way of saying, different columns. You see date, time of day and happiness rating, all the three columns we have. You will notice these three other ones that are italicized. These come with any default data set in Tableau. Don't worry about it for now. We'll go over that in a different video. Now everything on this right side of the screen is really a visual actually happens. Columns and rows just represents what columns are going to be within a column in both sections. You also have the ability to drag and drop different columns into each of the sections here and in the Marks card over this area, the square here, lets you change the visualization itself. Here's an example. Let's drag and drop time of day onto, let's say columns. Then drag and drop happiness rating on two rows. There you go, you have a decent looking graph that shows what your overall happiness rating is for each of the time of days. What happens here if I double-click on something and it automatically filtered it. You can see the photo up here. What's cool about Tableau is, there is the ability to undo, which you can see up here by just clicking Control-Z or just clicking the Undo button and you can always redo. It's very similar to anyone who's had experience with Excel. Just Control-Z, Control-Y to go undo or redo accordingly. Now there's a lot of different ways you can show this and you can click on the Show" Me" button on the top line to show different types of visualizations, especially if you never use Tableau before, this is going to be very, very useful. Go on here, you can see text tables. Let's click on that. Now just gave us a text version. Go back on horizontal bars is very similar to we have except everything is switched. One thing to know is we do click on one of these, "Show Me" buttons. Notice how the hills change. On the bars you have a column up here for happiness rating and then over here for the texts we have this on a text is technically is the text. You could change the type here. Now shown square and this is an example of a visual that just doesn't make sense. There's no reason we want to have both texts in the square behind it so this is something you wouldn't use. Let's control those either. Just try out different things. See what looks good, see what doesn't look good and you can drag and drop a bunch of different things as well. Now, let's go back to the bar. Actually unlike the other bar a little better, easier to see. There's a button that allows you to swap rows and columns, up here or even do it in the actual pill itself. This is what I'm saying when I say it's pretty intuitive because of the drag and drop features. Now one thing I want to show you is that these numbers are way higher than 10, despite the fact that inner happiness rating scale if between one and 10. The reason this is happening is you'll notice you see a sum of happiness rating. In other words, it's not taking the average for every afternoon. It's summing all the afternoons up and the rating is way higher than 10. You can do is you would click on this drop-down menu or just hit right-click and when you see measure, change it to average. Now that looks a whole lot better. This shows what each of the different averages are. Let's say we want to change, add some current to the time of day. You can drag and drop time of day on the color and you can see the change here. Let's say we also want to see some labels, drag and drop maybe on time of day. You can see what happens. They literally shows what the time of day was for every time of day, which doesn't make sense. We can take it off and take to it off you can note it as drag and drop it out of the way. You can see this red X next to the mouse. That means, okay, we're going to get rid of this. But it could make sense as few drag-and-drop happiness rating. On the labels and is doing the same thing mirror, it's gonna default to sum. Those few things you can do here. You can change the default properties and changing the aggregation to be average. Now I need to drag-and-drop. It's going to automatically show the average. I actually don't like how many decimals are here. Let's right click on the text and change the format. Now in this case actually, there's a lot of different ways you can format and I don't want to go and formatting too much. Let's actually take a step back and instead of changing the format here, let's change the default properties number format, and let's change it to number custom with one decimal place. Now I think that looks a little bit better and this is a decent graph. To go back to our framework, the first step is to explore and that's really where the most of our time for this entire side is going to be on. Because once you have explored the data and once you have a few visuals set up, the rest of the steps just fall in place. Let's keep going. Let's try out a few different visuals. Now, since this is one that we might end up using, let's keep it. You can rename the sheet on the bottom so let's call this time of day comparisons. Then you can do is you can just make a new graph or a new visual, I should say. It's like trying different things out. Now, one other big thing to know is when you drag-and-drop date, it's going to automatically show year. The few ways you can change this, for us, since everything is 2017, is not particularly useful to show date, but you can click "Plus" to draw down, all the way down to the day, but I don't like the way this is set up. I'm just going to undo this a few times, and instead of dragging and dropping date with the left click, I'm going to hold right-click and drag and drop date. Usually you want to do something green. Green means continuous, blue means discrete. Don't really worry too much about this yet. Let's say you want to see month, let's just stick with the green. Now, you have month on the x-axis. Let's do the same thing of happiness rating. Now you can see how overall happiness rating is changing over time. I like bars, you can scroll up to the marks section and you can change different types into area, line, I'm going to keep it as bars for now. I don't like the size, the way to it, just click on "Size" and change it. It's automatically changing the width of date. I like manual, this way I can change how big or small. Let's go with that. You can see here that the overall happiness rating is increasing over time, this it looks like it. Particular, a better understanding, you can literally just right-click hit "Trend Lines", "Show Trend Lines" and since it's going up that's good. That means in general, we are increasing our happiness. I would consider this useful as well. Let's rename this and call it monthly happiness. Now, we already have two different graphs that could be relevant. Let's keep going, and that's literally just drag and drop things, let's drop that here. Drop time of day, drop day, let's look by weekday. This is cool. It's actually gives a calendar type of view. Not quite, but could be useful. I don't like text. I think the text itself, it's hard to really understand exactly what's happening. I think it's much easier to see color. Let's see how we can change. Let's click on this "Highlight Table", there you go. You can see here, it's a little easier not to know what specific days in what specific time of days are the best for me. You can see here that Saturdays and Sundays seem to be the [inaudible] that's probably round were everyone's happy for obvious reasons. For some reason on Tuesday afternoon not feeling too well. You can see that here. Now you can change the color, click on "Color", "Edit a Color". Maybe you want to see some red-green comparison hit "Apply", and you'll see more of a stoplight type of setup here. Continue exploring a few different visuals. Continue exploring the data set, until you have a number of visuals you feel pretty comfortable using for your dashboard. Once you're comfortable with the visuals you have, it's time to move on to the visualization stage, which is simply setting up a dashboard. If you scroll the bottom, this middle button is a new dashboard. Click that button and you'll see the dashboard creation screen. Now a dashboard is simply dragging and dropping different visuals. It just allows you to set up a dash, however you want. In other words, you can drag and drop this graph here, drag and drop another one here. You can pretty much is play around with whatever you think makes sense. Now, I'm going to Control Z all these and show two different ways of showing data. What you just saw was the tiled screen. You see those squares that's how, you know, it's tiled. It's saying, this dash or this visual fits in this dashed through this tile. I personally don't like that. I would much rather use the floating option, which you can do by holding the Shift button and then drag and drop. You see there is no square populating, its just a floating square above the entire widescreen. What this lets you do, it gives you a little bit more flexibility to choose where exactly every dash and every graph is going to show. Let's put time of day comparison here. Maybe you want actually just put monthly and time of day at the top. Both of them are probably more like high-level things. Make sense to have them at the top. I don't want to show this color very obvious where each color represents here. I'm going to get rid of that legend. Let's resize this. You can just play around each of these. If something is showing that you don't want, just hit the "X" button. The other thing I want to mention is whenever you click on one of these visuals, click on the drop-down menu and hit many "Fit", it changes the way this fits within the screen. You can see here those is horizontal bar. I don't like this. I would much rather be able to see the entire view. Click on "Fit Entire View" and maybe it is expanded a little bit. That's pretty decent. Now continue this and eventually you can have a decent looking dashboard that can give you quite a few insights. Last thing I want to mention is the ability to filter. I don't think we started with filter let's go to the sheet. Let's drag and drop day and click, "Month", that's right click, drag and drop. Click "Month and Year". Let's say, artificial January and was not particularly used for on this, tried to do this. Let's actually try a different tribe. Sure, let's do this one. Same thing, multi-year. Let's just do generic. [inaudible] , you'll see the numbers change because it's going for my entire data set and suggest January. Now you can do is click on the table you just added the filter to. Right-click on or click the drop-down, hit "Filters" and then add the filter you just added, which was Month, Year of day. Then you see this popular. Now, what's cool is if you click the drop-down on this filter applied to Worksheets, all using this data source. Now this is going to finish everything at once. Now you can see everything just for May. If you want, you can have a fly only the Worksheet you care about. For me, I didn't want to filter for month since this shows month anyway. I'm going to take off the monthly happiness as an option. There you go. Now you can see this affects the left two, and this one is stay intact. When you're done with the visualized stage, you should have a dashboard created back in allow you that hope in life data. We already have some findings from the exploration stage back and write down, such as March 12th as pretty high rating. Overall, I am increasing happiness, which is good. Tuesday afternoon is pretty low, which is something that I might want to take into account moving for and maybe I can try to be happy with that day. You can continue to dig into whatever visualization you have. If you have interactivity may be clicking the things. Here's something interesting. On 6, 19 in 620, I have three low happiness relative to everything else. Engine. Maybe I check my calendar and see what happened those days were caused that low happiness rating. Anyway, do this for your visualization, figure out a few findings and then the last step is a storytelling. Really, there's not too much to say in terms of the storytelling since you are the audience. But use this to answer your question. Look at your findings. See what the lowest incidences and highest instances of happiness and try to understand what caused those. If you can maybe moving forward, you can be more real about it. You can do things that will increase your overall happiness. The core thing is as you continue to track your happiness, you can really see if what you are doing is working. Let's say you have a strategy, try it out for a month and your happiness does not increase. May be your suspicious strategy are blow but, that's it guys. When you're done with your budget, you'll save it on the Tableau Public and feel free to share what you have under the Course Description Project section. Continue at Tableau Public is everything is open source, which means you have the ability to download your books where other people have books, to see how they did things. Let's make this interactive, let's make this collaborative. Maybe you can learn from each other's findings.
8. Final Thoughts: That's it for this class, guys. I hope you gained value out of this, whether that's through the project you built, or just getting a better understanding of how data fits from the start to the end. One thing I will say is we barely even scratched the surface on this entire thing. Each of these steps can actually have an entire career dedicated to it. What I want to do is I want to get feedback from you guys in terms of what you want the next class to be focused on. Maybe you really like doing the dashboards. We can have another video focused on more advanced Tableau dashboard creation. Or maybe you want to get a better understanding of how to pour data and store it in a database. We can have an entire course on database concepts and using some sequel to move that data around. Thank you so much for joining this class, guys. I really do appreciate it and I hope you continue to learn more about this topic. This is the first day I've ever [inaudible]. If you guys don't mind putting some feedback down or maybe getting over you and be honest with it, I don't want to say it's amazing if it wasn't amazing, and if there was anything in particular you thought I can work on, please let me know that. I would really appreciate that for moving forward. I hope to see you soon.
Tyler Pernes, Data Analyst & Engineer
