 Right. Good morning. Good afternoon. Good evening, everyone. Welcome to the day five of the DHS to analytic tools online Academy. So we are kind of reaching the midpoint in this entire Academy which goes for two weeks. And I hope you, you must have learned a lot during these four days and today is going to be another kind of a hectic day because we tried to cover a lot of topics. So, that's why we decided to start right on time. So, let me start the presentation today. We don't have all the participants but given how tight today's schedule is I think we can start right so today we'll be covering this topic, the data visualizer. And some of the content at least the layout and where to access this particular tool has already been covered in yesterday's session when we were doing the table. But first of all, let us go through the objectives to see what we'll be covering during this day. Right, so there are several parts in the entire session that we'll be covering. So, let's see, let's go through one by one. What we will be doing is we will be discussing on the overview of the data visualizer interface. And then we will also discuss about how to select the inputs for a chart. So we'll be discussing how to add the data organization units and the periods dimension. And then we will discuss around about how to arrange the layout of my chart using categories series and filter. So these three terms are totally new to you. So we will discuss what these three terms mean. Next in the part two we will be discussing about how to work with this aggregations in my chart, and also how to add additional group selections, such as you may you must be already familiar about concepts such as organization groups into the chart. And then also we will discuss around future options like how to add titles, how to add 100% stack value sort order and target lines. And finally about how to download a particular chart. And in the part three we will be discussing around the topics of pie charts, radar and spider charts as well as gate charts. And then finally, as the part four we'll be discussing about cumulative values options and year over year chart time. So, I know it's a lot of topics that we are hoping to cover in this, I mean during one day but we'll have to do it so that, I mean, we'll be able to start more to the next session without dragging few of the contents from today's session to Monday, right. So, let me share my screen again. I'm seeing a few participants are joining now. Right. I'm sharing my screen again. And I'm in this instance, our exercise instance, the same instance that all of you have access. And right now I'm in this very familiar interface, which is a dashboard. So, you already know how to access the data visualize application, because we also did the same yesterday when we were doing the pivot tables. So let me click on the data visualizer, what we see here, right, and it opens the data visualizer application. So, first of all, just to recap, what we did yesterday about the components of this data visualize interface so on to your left side you are seeing the dimensions so here we this is where we select the data, the period and the organization dimensions the three W's of PHS to. And then, if we have in addition any other dimensions that we have configured this will be visible to you, right. And on right hand side in the top, we have like some selections series category and filter we will discuss about them a bit later, and this is where we are getting the final visualization out. So, let's do something I'm going to start by clicking on file and let's open one favorite item which has already been saved. So I'm going to search for this item that we have saved previously called HIV cascade. So I'm going to search HIV cascade and I get, and I see this item here. HIV cascade, and you can see here the item type which is a column chart, right, I click and open this chart. Right, so what do we see here. So, previously in fact yesterday what we saw in this data visualizer in this visualization component were all tables, but now we are seeing charts. So entire day today we'll be seeing different types of charts, and we will talk about what these chart types are one by one a bit later. But if we just concentrate on what we see here, what is this is I think all of you must be familiar. So what is it, if anyone can answer, just to see if everybody's listening. What do we see here, anybody. That's the simplest question I can ask for the day. What do we see here. It's a graph. Data. So we have people living with HIV, and on the X axis, and then these are people in training land last 12 months. The positive cases, those on ART and those retained in the in the on ART in the last 12 months. Okay, great. Thanks. A graph column graph column. That's yeah exactly so what I was actually expecting for you to start this like it's a column column chart that you're seeing here, and you nicely describe what we have in X and y axis. We are seeing some data, right. Three types of data that we see HIV positives people doing it there and we are people doing it. People leaving HIV returned for the last 12 months. And on the Y axis we are seeing the values, the data value. We have the data only for the last 12 months for the entire training land. So that's what we see in an action. So, again, let me just try to hover over what we see here. So when I, I will, I will, I have not yet discussed about what series categories and filters mean but let's see what we actually have here. So in the period we are seeing the last 12 months. In the filter and here we are seeing of course the organization unit which is the training land and when I hover on to this category is the data dimension. We are seeing like there are three data items that that of course what you are seeing as individual columns here. Right. Okay, so let us try to edit one by one. So I'm going to click on this period dimension here. And now I'm seeing the relative periods, the available relative periods for the months category is mentioned here and the selected one is 12. Right. Okay. So let's try to change that a bit. And let me select last three months so I can double click and it goes here. Let me also remove the last 12 months because like this of course is something that I mean it doesn't make any sense to put both of them there. So I just removed last 12 months and what I did was I put last three months and now what I can do is I can click on update. Let's see what happens. Right. Now we see that the values have gone down, right, the number of the final value, the figures, and we are also seeing that data that you see in this visualization is only for the last three months. Okay, so I'm trying to recap what you did yesterday about pivot table so this there is nothing new here. Okay. And now let us do something like this. Now here we see organization unit appearing under series and the data appearing under category. Right. We have still not defined what series and categories are but let's let's try to, you know, like shuffle them so that now the data goes to the series and for the category we get organization unit. Right. And I click on update. So what do we see here. We see that basically what actually changed previously we had all these I mean individual columns in one color but now we are seeing them in different colors right we are we are seeing the people living with HIV positive in green and blue is the one ARP and red is retained on ARP in the last 12 months. Okay. Right. So this way what we understand is like in when we just shuffle this individual dimensions around series category and filter, whatever the output that we that we plan to have here is going to change. I hope everything is clear up to this point. Okay, so let's try to change it one more time. Let's put a period here. Okay. And let's try to get the data here and organization units I'm going to put to the filter and I click on update. Now you're getting a total different visualization where you see that whatever we saw here before has been grouped into separate columns. Okay, so let's just focus on what we see next and why you access. So here in the x axis we are seeing the data items right HIV test positives PL HIV new one ARP and in last 12 months. And again in the y axis we are seeing as how we started before the figures, but the difference that we are seeing now is the grouping the original grouping that we have in x axis has have been further subdivided into colors. Now this each of these color bars represent the months. So this way, what we can do is we can compare each of these values, the data items across the period dimension. Right, so that's what has taken what has actually happened. Right. Okay. Fine. So, now that we have seen some modifications that we can do by changing or shuffling each of these dimensions around this series categories and filter. Let us actually have a look at what the series categories and filter me. Right. To do that, I will move on to this presentation. So what do we see here. So the visualization is not too much different to what we saw before, right, what you are seeing here is a column chart in the first place, right. But then again in in that column chart we have identified some different area. So we have numbered them as one two and three. Okay. The number one here, right, we are trying to show a number one and two both we are we are trying to focus on something that is appearing in x axis, right. And then number three here, we are talking about something that is affecting the entire visualization. So let us take one by one. So the first thing, which is of course what we refer to as categories. Okay, number one. Every what we see here is a period dimension and you can see here November 2018 December 2018 and January 2019 here. Okay, so now in this type of chart. Now remember what I mean what about I mentioned just now. These definitions that I'm going to tell now about categories series, especially these change depending on what type of chart you are using. So here in this column chart. The categories will always be the x axis, right, as you can see here, the x axis. And we can see the period is the x axis in this example, you can see November December and January. And when we first open the chart, the data data was in the category and the data items were listed along the x axis, if you can remember. And y axis in this type of chart is always going to be in this, with this configuration, the values, right, you can see like now 2500, 5000, 7500, so these are always going to be values. Right, so basically what you can, what you have to realize is that in this type of column charts, it's the category is going to be next axis, and it's where it's basically the groupings of the data that we are showing. Okay, so if that is so what do we mean by the series. Right, so what series does says, these in fact, further subdivide what you are seeing on the category dimension. Okay. So here, you see the data in the category which is like I mean the periods November December and January, and what series does says like whatever the data that is there in the, in each of these dimensions so for example the November, it has further divided that data. So that is why we are seeing HIV dispositives people living with ART and the last 12 months under this November 2018 category. So the same thing we are seeing in for December as well, and also for January. So what you have to understand is like category in a nutshell is going to be next axis for these type of charts, right, and series is going to further subdivide or subgroup the data that you are seeing in the x axis defined by the category. I hope it is clear up to this point about categories and series. So if that is the case. What do we mean by filter. So, the filter per se like just as how we understand the concept of filter in layman's term, it's it kind of filters out what is going to appear in this visualization. Okay, so basically, like, before we decide on series and category everything that is going to appear in this visualization has been defined by the filter. All the data that we are going to show in the visualization is getting filtered by the filter that we define. And once the data is filtered that data we are going to arrange the layout for categories and series. Right, so now here as you can see the organization unit has been defined in the filter as training land. Right. So, because it has been decided decided that training land is going to be the highest point at which we filter. The data that you see in this visualization is going to represent the entire training land. So in case if you can change. If you can filter out the visualization for a given region say animal region, the data that you're going to see here is only going to be representing the animal region. Okay, is that clear any questions up to this point about series categories and filter. Now, before you ask. Say like, can you define the category as something that is going to be x axis in general across all chart types. The answer is no. Right. So for example, there are some chart types where we can't define x and y axis. So for example, in pie charts we don't have x y axis and in bar chart sometimes the axis changes right so so I will talk about it in time to come. But for now let's focus only on column charts and try to understand the concepts of series categories and filter. Any questions up to this point, not. We can decide we can discuss about individual chart types. I think now we have most of the participants joined. So just to recap what we have done so far is like we discussed the objectives for the day. Basically we'll be covering all about the data visualizer. And then we discussed the interface of the data visualizer app, which even though you joined late, you must be familiar because it's the same as what we used yesterday for pivot table. And then we discuss about this dimensions and how to configure them in data visualizer so the important three, three concepts that we did that that is that we did not discuss yesterday. So are these these new the terms series categories and field. So filter basically defines the scope of the visualization so whatever we we try to visualize in this graph. The filter defines at the top the boundary. Okay, so everything gets filtered out from what we defined here and appear in the visualization. So column charts category basically define the x axis, and whatever appearing in the x axis is further subdivided and grouped by the series in this type of column charts. Okay. So if that is clear, let me again stop sharing presentation. And I will try to discuss about something else. Okay. Right. So I'm going back to the our demo instance, the exercise instance, and here we see the series categories and filter and and then we are also seeing something over here, which we did not focus yesterday or chart time. Okay. So here we are seeing now yesterday we were using pivot table and now today we have, we can discuss so many other different chart types right we have we can see like most of you must be familiar column stack column bar stack bar line by your area and there are so many chart types. Right. So we can use any of them for different visualizations, but there are some generally accepted concepts on which tab with chart type suits which particular scenario. Right. So what we are going to do next is to just go through a brief overview of which type of chart type to use. So to do that. So let's go on to a different presentation. Right. So this presentation, as of now will not appear in your the learning management system so don't panic this will be uploaded into it once once I conclude this session. Right. So, don't worry because it is not showing in your elements. Okay. So let us see what are the types available and when we can use it. I hope you can see this screen. Now, the thing is this is a very overwhelming crowded slides so pardon me for that, which highlights the different types of visualizations that you can use as charts. Right. So the, the concept I'm not going to discuss in detail of what you are seeing here but what I want to highlight is like the main question that we ask is, what would you like to show. Right. So, when you are like always think before you design a visualization what you actually expect to show and to whom are we showing. So for example, when we are deciding on what to show, we have to decide are we trying to show a comparison, or we want to show a distribution, or it is easy to about composition, or about relationships. Right. So, just give me a minute. I'm having some issues with zoom. Great. So, thing is we are okay now I think you must have been looking at what appears here. I'm not going to go through all of them. And I must also mention like some of the graphs or chart types that you see in this particular slide are not available in DHS. So for example, what we have to think is, say like, if I want to show a comparison, right, if I'm going to do that, I just have to decide, is it among few items, or is this comparison over time. So if it is going to be comparison over time, the probable chart times that the types that we can use are mostly line charts, or maybe even column charts, or we can use area charts. And then again, like, which one to choose across these options, we can, I mean, decide on a few other minor criteria like how many periods, things like that. And if it is going to be among items, then we mostly will be using column charts and bar charts, right. And if you are going to use relationships, we can use scatter chart. Unfortunately, we are not discussing around about scatter chart because we are mostly focusing on 2.35, but that chart type is available in the DHS to version 2.36, which will be not which we are not covering during this economy but if you're interested you can check the latest version of DHS to and it's suppose scatter chart. If you want to show distribution, there are few chart types available, right. And then, again about composition. There are chart types such as column charts, pie charts, right, that that we that are available in DHS to, and we can decide which chart type is most suited for a given visualization that we are trying to show. Okay, is that clear, you don't have to memorize anything this I just mentioned as reference. Okay, so now let us go through one by one in and compare each of the chart types, right. So the first chart type that will be most of the time dealing with is called column charts, right. So, can someone tell me when we are going to use column charts. In the simplest use cases, we can use column charts, like when, I mean what type of scenarios matches the use of column charts. Anybody, yes Steve. When we're comparing data, looking at different values for maybe as this one is like different categories or different groups of people, we can use the values that we are looking at for the data elements or variables of interest. We can use column charts for that. Right. Yeah, that's great. Any any any further answers. Comparison between like, in this instance, between districts how each district has performed. Exactly. Yes. So, basically, column charts is a nice type of visualizations if you want to show comparisons right. For example, there are a few things that we have to keep in mind when we are using column charts. So, the thing is like comparison if like I mean all these things that I mentioned here. None of them are hard and fast rules right there, they are there are exceptions and there are different norms accepted by different entities, but like these are the ones that are generally recommended so right. The number of categories is quite small. Right, you can use column charts, right. And if usually if the time dimension is is something that we are concerned about we generally put the time dimension into the x axis or the horizontal axis. Right. And then if there are trends, you can use column charts but may but but I will explain about it in the next few slides. So, column charts is not the best chart types for us to use if you are talking about trends, but if the number of data points are less, we can still use column charts. Okay, right. What do we see here, this is called stacked column charts. So when can we use stack column charts. Anybody. We need to do this aggregation. Disaggregated categories. For example, we are seeing here in this, in this particular example, we have the male and female, basically gender desegregation, and we are using stacked column charts. So here, basically we are talking about the composition, right, so the data is there in one column, and we can talk more about the composition of the data that is there in that particular column. Right, so for example, we can see the blue and green components together consist of this entire column or entire bar right so to discuss about the visualizations we can use. And then we have to be mindful not to use too many compositions at a time. Right, because like you can think of a column here in like all the rainbow colors and then what happens is, our human eyes are not good at, you know, comparing across a particular column. Right, when the chart when the categories or the composition items become too much. It will be very difficult for the, for the person who sees to compare. Okay. And then also make sure the composing parts are relatively similar in size right but like if these are very small and like some are very large then again it will be very difficult to compare. So let's move on to the next chart type which is the bar chart. So, right now, now this is somewhat similar right so we discussed about column charts before, and this is again a category of a column chart it's just that the appearance is somewhat different. So what do you what what can you say about bar charts when can you use bar charts. What would be the situation that you can consider using a bar chart as opposed to the column charts that we showed before. And the range of products are. So Richard and I could not hear you. And the range of products. It was not quite clear but yeah. What do we need both comparison and composition. Oh, yeah, but why don't we just use the column I mean can't we do it by using column charts also. Yeah, I mean it's a bit of a tricky question like why don't we use column charts and I mean, you know for this example. Yeah, why is it because of the values, if the values are large. That's the thing right so the thing is like say for example if the number of categories are like because if you can remember in when I was explaining about columns if the categories are like limited you can use this the vertical columns right but if the categories are larger. So we can have a better visualization if we arrange it as bar charts right and also when like for example if we have this longer names. I mean, this is another like reason why some people use bar charts like when when the names are longer. And if you want them to appear here in our normal vertical charts if you want to appear them in the next access that won't appear that well so for that also some people use bar charts. So the next chart type is line charts. So when can we ideally use line charts. For visualization of time series data visualization of time series data yes that's that's a classic example and when exactly like there are many use cases of visualization of time series data but what are we actually trying to do with line charts. I identify trends. Excuse me. Yes. Maybe back a bit on the bar chart please. I want just to ask one question. Yeah, please. Sorry. Just scroll back to the previous slide. Previous. Yes. So, for the first point where it lighten that number of actually we use bar charts when we are using when we have a number of categories. I was wondering if the statement is well lighten or is it the number of categories which should be greater than seven so that we can use this bar chart or it's these are the values of categories which should be greater which should be higher so that we can use this. The number of categories of the values of categories. All right, so, as I mentioned before now these are not hard and fast rules, and these are just general accepted concepts, because like what I simply want to. I would say like in categories that are appearing in the horizontal bar, horizontal column, the column charts, it wouldn't that nice, it wouldn't be nice. But in bar chart what happens is because the x, the y axis is exactly where these categories are appearing and have slightly larger number of categories that we can accommodate. There is no cut off limit like if it is seven, eight or 10, nothing like that and even you can still if you think like for my visualization, I think that this chart types looks much better and it is justifiable. You can go ahead and use it. I mean there are no particular cut off values it's just that I have highlighted because like, otherwise you would ask the usual question that comes is if you don't. As I just mentioned, if too much you will always ask what would be a number that we can decide so that's why we have mentioned some numbers here but I mean it's totally up to you, but the general general concept that I want to highlight is the y axis we can accommodate more categories, compared to the x axis. So that's where people are deciding to use bar charts. Okay, thank you. So trends, sorry, coming back to line charts. Yeah, so I can remember who mentioned but the, I mean one important thing is like line charts are the best suited if you want to do some trend based visualizations. And again, the data visualization across the cross period of time, right. And if you can remember, I mentioned, even for column charts we can use for comparison over time. But the thing is if the number of data points are less you can use the column charts, but if the number of data points are very, very high can always use the line charts. Right, so this is one major application of major chart types that we usually use in DHS when we want to highlight time based friends of data. Okay. Area charts. So, like, what can you say about area charts, when do we use it. Basically area charts is again another variant of the line charts. It's just that it's kind of giving more focus on the area that is below the line. Right, so you have the line chart right here. And it's also, it has applied some color to the area that is lying beneath the particular line that we are concerned, right. Basically, these type of charts, right, because because it is more concerned about the area it's best suited if you are trying to present the cumulative values and the change of cumulative values over time. Right, so for example, if you want to talk about say a number of COVID-19 cases of a country across last so many months, right, and you want to show cumulative value, you can use this area charts. Right, so that's one application of when we can use area charts. Right, then we have another variant of area charts called stacked area charts right. Like what, when, when would you apply stacked area charts as opposed to just a simple area charts. What can we show more in this type of chart. It's again similar to the column charts right here we can show the composition right. And then again, composition plus if you are talking about cumulative values you can use the stacked area charts. Okay. So next we move on to a very common chart called pie charts. When are we using pie charts. What's the classic use of pie chart I mean as opposed to any other chart. Comparing components. When comparing. Components. Components. Yeah, but I mean like, what is the most important thing about a pie chart, which you can't actually I mean, which has to be there, which was not a concern when we were applying the column charts distribution. Distribution. Yeah, so what about the distribution. Total value has 100%. Exactly, I mean that was the answer I was expecting so here like now, now the thing is for like, whatever we are trying to do right we have the entire circle which is a 360 degrees okay so we just can't fill just one section. Okay, right or one wedge of the circle and just not think about the other right so that means whatever we are trying to highlight here has to be a part of the entire fall. Okay, so that's exactly what Chintaka mentioned so he mentioned that it has to ideally all this should that add up to 100% or else, all this should that up to 360 degrees that that formulate the circle. Okay, so that's the most important thing. So that is why this is best used when we want to visualize a part to whole relationship or a composition, right, because we can use stack charts to show but here there is always this part to hold the proportion concept is always there. And the other important thing is, now this is not meant to compare individual sections to each other or to represent exact values, right, that we have to be mindful that we are not going to you know like compare individual sections when we are using pie charts you can say yeah we can, obviously right here you can see now looking at this one we can see female is greater than male. But this is not a concept that we can apply in a generic way and try to you know like represent exactly so why I say so. Okay, let me move I'll show you another example. So what do you think about this pie charts. Any comments. I mean, vivid, beautiful, so many colors, any comments. Is this a good pie chart. It is a five category. It's a good pie chart. Okay, any other comments. Because it's okay so it's not, it's not presenting the data percentage of data. So it's not a good pie chart representation. Okay, right so you are saying it is not a good pie chart because it is not showing the percentage of data right okay so if what if I say right now here it doesn't appear but I can mention the percentage or the actual numbers in each of these ones then does it become a good pie chart. So for example, I mentioned the values the exact values as a text in each of these wedges. Does that make it a good pie chart what do you think it will be good pie chart if it's if the distribution equal to 100%. The distribution will come up to 100% for sure. Yes, it's a good and it's a good good pie chart. Okay, any other comments. Thank you for this. No 된 so no no. Okay no legends. Yeah, the thing is this right so you must have seen. I think at least in my country like most of the media when they want to highlight you something that they want you to think they do these kind of tricks right so the thing is, I would like what we have to like we can justify our visualization saying okay we show this we are having a text and things like that but what we have to be always mindful is what we are showing now right whatever that you are trying to visualize is what catches the mind of the inducer first right so they will go into specifics about reading the the the legend values or the text values inside this each of the edges if they have time otherwise they will just glance at it and try to grasp whatever they they perceive right so in that sense these charts are not ideal because if you can remember let me show you the next I mean something different okay oh what do you think did you see this aspect before now here you are seeing a set of pie charts right and for the same scenario beneath here you are seeing column charts what are the differences like what do you get by looking at these two chart types and then with this can you say something now if you go back to my initial question whether this was a good pie chart or the whether I mean whether this was a very good scenario of application of pie chart if you can elaborate now what do you think in a pie chart we are not easily able to assess the magnitude of each category here now here in the in the bar chart below we are easily able to understand which is maximum and which is minimum whereas in the in all the three pie charts they look more or less the same exactly yeah that's the answer so the thing is like I mean that's what I said like you you can justify okay we can prove I mean yeah true because the number of categories are around five yeah we can use the pie chart no but the thing is if our aim of the visualization is to show the composition or a kind of a comparison and we are seeing that each of the wedges in the pie chart is going to be more or less same then the human eye I mean we really fail to differentiate or compare the values by looking at a pie chart as you can see here now here if you use a column chart we can nicely see the comparison but the pie chart fails at doing it okay so that's what we have to understand okay right so going back to pie chart and and about when to use it so make sure that it equals 100% right and it has to be less than six categories I mean otherwise it's going to be very difficult but always think what you are trying to show right so for example if there's a clear winner right I mean if you have a very large wedge that you are that you can show in the pie chart you can use it but otherwise if it is similar to the visualization I showed in the previous slide please don't use it right and ideally it should be a simple like for example two category pie chart where which which can easily you know highlight the idea that you want to highlight the end user okay and this was what I highlighted in the previous one if these are identical please don't use it okay yeah what is this this is called gate chart and in dhs2 we have gate chart when can we use gate chart right so I will answer myself because it's so so simple usually now this gate charts are ideal if you want to highlight or focus about the single key right and usually even in dhs2 what we can do is in addition to highlighting a figure right you can see this is 81 we can change the color right with a legend so that you can actually give some idea in addition say for example now this 81 may not mean anything we may feel okay 81 is good but what if 85 so for example if if our country decide on a legend where we think 90 or above is the best right and maybe 90 to 70 is average or anything below 80 70 we don't even care it's very bad right but then nobody will understand this concept if we can't apply a legend because they will think okay 81 is it's not bad it's maybe even good but it's not so so the thing is there are two concepts you can here of course highlight the value and you can apply a legend you can weigh that second concept that I want to highlight which is the legend okay so these are ideal for examples such as if you want to show the progress or if you want to concentrate about key performance indicators or single measure right and the most important thing is you can quickly look at this and understand the idea right right so I will stop the different chart types here we'll be talking about few additional ones about radar chart and single value charts which I will discuss when I'm doing that particular session but I just wanted to highlight and compare the few important chart types that we are usually dealing with when we are talking about the visualizations in the dhs2 data visualization okay so let me go back to excuse me yes yes questions yes back to the pie chart yeah we have seen that it's a face to use the pie chart when at least you have two or less than two categories to compare less than two of course no I mean less than two means one so one category yeah yeah I mean at least two yeah sorry did I stand to the presentation yes so does it mean that more than two it's not divisible no not like that what I wanted to say was ideal if we have two categories because then we we actually as I mentioned before we want to show a comparison which attributes the entire whole so the thing is like if it is something like male and female just by looking at the pie chart you get the idea but whenever you have more than two categories what we always indirectly ask the user to do is compare compare the sizes or the size of the wages this we give to the user right so if we want it again depends on the type of user if our end users are not too tech savvy or like they are just ordinary healthcare workers and we are meaning to have these visualizations in a dashboard available to the field health staff then to confuse them with the I mean large categories in pie charts where they have to you know compare by looking at the visualization may not be the best method that's what I wanted to highlight so it's now you may you may realize like it all depends on what you want to do there are best practices recommended practices you can always deviate from them if you can just okay in otherwise I was wondering if there's any way to represent these kind of institutions in terms of percentage without using without using pie charts I'm just I was just wondering if there's any way apart from pie chart but I have I have now I understand what you mean yeah there are ways I mean for example when we were doing the pivot table we talked about applying legends right so yeah invariably we can I mean by using legends we can convert it to a scorecard so when we do that scorecard and things like that maybe another type of visualization where you compare the colors as well as the figures are already there so yeah there are different ways of doing sure okay right fine um right so we we can do the uh no I mean I think we I will cover the next component also and then we go to the ungraded exercise okay uh juneid any questions oh uh sorry juneid he's not clear uh is there I mean you want to ask a question I think it's not about a question so let me mute you right okay so let me share my screen again fine right so let's try to draw a new chart so whenever we want to new draw a new chart what we can do is if you are inside the data visualizer we can click on file new and it'll always ask whether you want to save it right or like you want to leave right so I will just click on yes leave so I don't want to save it okay right and what I'm planning to do is to draw a chart about tb notification right and to see whether there has been any significant changes of tb notifications across time right so usually um like I want to do a kind of a comparison um by drawing this tb notifications right and let me ask a quick question in tb programs what's the usual frequency of reporting like what's the usual period type uh free period frequency they report data in tb in tb most of the time it's on cut their basis exactly it's it's a usually quarterly right okay right so having that in mind right if we want to compare right friends of tb reporting what would be the chart type that you would like to use what would be the most useful to compare the friends in which period of time uh sorry a period of time it could be quarterly we want to compare the reporting rate yeah among different yeah okay let me explain the scenario a bit a bit further right so for example um I would like to compare with the now there are different tb notifications right we have all cases we can compare new and relapses we can compare new and pulmonary bacteriological confirmed right uh or power clinically diagnosed so we have different data items right that that we report in tb tb program right so we want to check how this reporting of different data items have changed across a period of time maybe like uh uh uh reporting over three years something like that or two years so if that is what we are going to visualize what would be the chart type that we would like to use you are over your line chart okay right so yeah okay yes exactly so there were two answers the first answer I would for now I would just pretend that I did not hear that because it might confuse many of you so I will not talk about this year over the year line chart just now but yeah I just wanted to know broadly what type so we have different types like columns lines five so that way I got one answer saying line charts everyone agree okay there are no answers yes okay then let's use line charts because as you can remember I mentioned like the line charts would be the ideal if you are using um it for comparisons right to show the trends right I hope you can see my screen let me share again I had some issue okay right so I have selected line charts here okay right and then what I'm going to do next is I have to decide on the dimensions so first of all the data dimension okay so let me click on data dimensions this I'm going a bit fast because you are already familiar with it so I will uh select I will keep the data type as indicators and we have different groups so I will take the group call eb case notification right and here I will try to type say notification notification rate okay right so here we have four data items selected like one for cases one for relapses bacteriological confirmed and pulmonary clinically diagnosed right so all these four items I will select onto this side okay so that is confirmed like now this data dimension is done I will now I have two options I can click on hide or I can click on update the difference is if I click on update it will actually update the visualization if I click on hide it will select and it will not visualize anything okay so I click on hide and keep it just like that so I can see when I have however my mouse point I can see that the four items have been selected but the visualization is not yet there right okay right so fine I will just keep it there and I have now now here we have now when I did that it has automatically gone with a series okay so the question I have is like based on what we learn the bar chart type example what effect will the data being in the series have on this line chart so if I repeat my question you can remember what I drew on the bar chart the column chart right so based on that how do you think I mean like applying data to line charts how do you think I mean where do you think this data dimension will represent in this final graph you can visualize right the final output and tell me where will these data dimensions appear in the line chart how will it be like will it be the x axis or is it something different when it is there in the series anybody I think x axis okay x axis is like this right here we will have some okay fine x axis any other answers any more answers right so before I mean so I will just stop it there I will just post it there and I will go further and apply something for a category dimension as well so let me click on the period so automatically period is in the category dimension so I click here right so here I will select quarters right and I have quarters for each of the years sorry now here I have only relative periods so let me do something I will just get rid of 12 months and I will go for fixed periods okay and what I'm going to do is I am going to have for the period of last three years all the quarters so I will select quarterly in year 20 I will start with year 2018 right so I have four quarters in 2018 and I put it here and I select 2019 then push it to this side and I select 2019 and push it up to this side is that clear what I did I selected fixed period as opposed to relative because I want to compare a long time not just one year right because this is quarterly later right so I selected last three years and I selected all the quarters in the last three years okay is that clear and I click on height done what about the org unit let's keep it at training land for the moment right fine so everything is set and all I have to do is so I have selected the chart type series is done case categories is okay filter is okay and I click on update right what will happen this is what I see okay so like let's try to address this I mean apply these concepts of series categories and filter so what has happened here now filter I mentioned to you is like what is sitting right at the top filtering everything that comes into this visualization so we have applied organization which is training land so what it says is all the data that is representing here is from the training land right and the x-axis here is basically represented by the category which are like what we are seeing the time periods so each of the quarters right and basically each of these lines that you see are reflected by the series dimension okay so in line charts what you have to understand is individual lines are reflected from the series dimension and you can nicely see we have so many data points like at least 12 of them and you can see the trend how these have been changing over last almost 12 quarters is that clear okay right so let's try to do something now here we have the data in series and categories it's a period okay so we can you know like toggle them so let me move data to categories and period to series right and let me do update and let's see what happens okay great this is what we get what do you think any comments is it a better visualization than the previous one or previous one was better what do you think wow yeah like two or three people were telling previous one was better right yeah now tell me why why you think previous one was better is our sensor should be on x-axis so when you put categories on x-axis it becomes some bugs huh categories on x-axis okay right so first of all now can we make anything out of what we have here in the first place like how many lines are there all together here how many how many lines are here yeah lines of all quarter exactly count of all the quarters right so now that means we have around 10 lines here and we actually can't figure out which one I mean which line is going in which direction so that itself is enough to justify that this is not a good visualization and in addition it would have still been better if we were expecting to note significant differences across these quarters like say for example now it all depends on what we want to highlight right so our primary focus is when I mentioned you the requirement was to compare how reporting rates have changed over time but like if I what I wanted to highlight was like across these so I mean like across these four data items there have been significant differences across different different periods and I wanted to focus on that and of course I have noted by looking at the data that there are significant gaps across the quarters then I could have used this kind of a visualization but I knew before start or like first time I did this or maybe when I looked at the pivot table I knew that there were like only slight differences across each of the periods and it was just I mean so they they are they are like mostly ranging between like two or three digit values right so by having that knowledge I should know that this kind of a chart is not going to work okay is that clear fine all right so I will just briefly stop here it's time is yeah almost what we have done it for one hour so what I want you to do is log into your learning management system open edX and do the activities one and two up to this point so we can take like 15 minutes for that and in that you can also get a take a brief bio break and maybe we can see you off around 220 so exactly in 20 minutes sorry exactly in 15 minutes we can stop is that clear you can do this ungraded exercise not the graded one up to part two yeah okay right so see you in 15 minutes if there are any questions please use stack my colleagues will be there also like they'll be helping yeah okay all right then see you in 15 minutes okay thank you