Category: Tableau

Tableau Heat Map Tutorial: 3 examples

Leave a comment

What is a Tableau Heat Map?

Heat map in Tableau is a data visualization type for which you need to have one or more dimensions and 1 or 2 measures. In these heat maps, Tableau is displaying a table consisting of many squares. Tableau can represent these squares in different sizes and colors. You can choose measures that will reveal data by size, color, or both.
 
Tableau heat map

How do you create a Heat map?

To create a heat map, you need to have a data set with at least in dimension and one measure. Ideally, you must have two dimensions, of which one should be a date and one or two measures. When you select your data source, follow these steps:

  1. Select your dimensions and measures when holding a CRTL button.
  2. Press Show Me button in the top-right corner and select Heat Maps to create our Tableau heat map.

What are the best practices to use heat maps?

To use a heat map in Tableau is a go-to practice when you have some measure that you can analyze by timeline or sub-categorical dimensions that divides the sum of that measure. If you have too many rows and columns in your heat map, then you can add filters to make finding insights easier.

Example 1. Determining which region has the most sales

 
We will use Superstore Sales data for this first example. Firstly, connect to our data by selecting the Orders sheet as our data source. Then we immediately go to our Tableau Sheet 1 and renaming it to what we want to find in it. In this instance, change Sheet 1 to Product Subcategory Sales by Region. As a result, we now see the modified sheet header and can go to the next steps that will create a heat map.
As you can see in the following video, we are selecting Product Subcategory, Region, and Sales. Then, we choose Heat Maps from Show Me list. As a result, we have our first heat map that can be formatted to look better. We switch column with header because instantly, Tableau put it in alphabetical order. Then, change Region header alignment to vertical.
 

Tweaking the heat map

Even though a Tableau heat map is intact, we still not have the best version of the chart in front of us. Indeed, there are a few things that we can improve here. First, increase the size of a heat map so it would look more appealing. Second, switch SUM(Sales) pill in the marks field from Size to Color and choose whatever color palette you like.

Now we can see that most sales were in Ontario, Prairie, and West regions. Also, it’s clear there were no Office Machines sales in the Nunavut region as it’s rectangle is white.
Moreover, we can use the second measure in the heat map to display both different sizes and colors for rectangles. Let’s change SUM(Sales) pill to size and drag Profit measure to colors field. As a result, we can understand which subcategories by region have the most sales (width of rectangles) and profit (color of boxes). With this addition, we can recognize negative profit cases. For example, we can see that Bookcases our store sold with negative profit in Yukon, Northwest Territories, Ontario, and Quebec.
 

Example 2. Best App Store games 2008-2019

Let’s open the 17k app store games database from Kaggle. Steps that I use to create a Tableau heat map are in this video:

Using various filters, you can determine multiple app store games that have the biggest User Rating Count. As we know that the best games get rated more, the big User Rating count can indicate the best seller games in the Apple app store. If you want to know the best Role-Playing games, filter this genre and set the minimum User Rating count such that optimal number games could appear in a worksheet. For example, after placing the minimum User Rating to 20.000, you can see a Tableau heat map consisting of probably the best Role Playing Games in App Store history. Tableau Public worksheet is shared below.

Example 3. CO2 Emissions By country in the Tableau Heat Map

This example is meant to show countries with the most significant CO2 emission and also changes in the emission numbers in the timeline. You can download the CO2 Emission data here. I am not going into further aspects of what steps I am taking to build a heat map in this case. Just watch the video below and digest it. 

Conclusion

To summarize, a Tableau heat map is an excellent alternative over traditional excel tables with numbers. In numerous circumstances, it is easier to find patterns using the size and color of cells over numbers.

  
Tableau , , ,

Tableau pie chart: full tutorial

Leave a comment

The pie chart concept comes from old times. The pie distribution problem: how to divide a pie into X equal parts for X people. Nowadays people use pie charts to asses what components have the biggest share in some totality. In Tableau visualization, a pie chart is one of the most popular charts. For business applications, this chart is most appropriate to demonstrate profit by country, sales by client or expenditures by a company branch.

business pie

For this tutorial I am going to use the same Superstore Sample data set that I’ve used in Tableau bar chart tutorial.

 

Creating a pie chart

To create a pie chart in Tableau, we need to have 1 and more dimensions and 1 or 2 measures. In this tutorial, I drag Sales to rows and Product Category to columns. However, a bar chart was created automatically at first. After that, we press Show Me and select pie charts. By doing so, we create a pie chart.

creating pie chart

Other creation method

If we want to avoid creating a bar chart, we need to follow this instruction:

  1. Select Pie in the Marks field;
  2. Drag Sales (a measure that you want to asses) to Size and Angle buttons;
  3. Drag Product Category (a dimension by which you want to divide your measure) to Color button.

different method pie chart

 

Formatting a Tableau pie chart

Changing size

Changing a size for a Tableau pie chart is one of the most annoying tasks. We have to move our pointer to the right next to a pie chart. As our pointer changes it’s appearance to a two way arrow, we can press, hold and drag it to the right to widen chart size limiter. Also do the same with the bottom limiter to attain a bigger chart.

changing pie chart size

Display labels

To display labels, drag Product Category dimension to a Label button. If you want your labels to be inside of a chart, you can click on those labels and drag to a center of an angle or wherever you want. To see what percentage of a whole share each category has, drag Sales to a label button. After that, select the SUM(Sales) label mark and add Quick table calculation – percent of total.

Tableau Donut chart

Donut chart is a variation of a Tableau pie chart. To create it, we need to:

  1. Drag Number of Record measure to rows column twice;
  2. Set measure to a minimum for both of those Number of Records marks that are in rows field;
  3. Then we select Dual-axis in the second Number of Records mark;
  4. In the marks field of MIN(Number of Records) 2, remove Product Category from colors;
  5. Then, press the labels button and mark out “Show mark labels” field;
  6. Select white color or whatever color that suits your Tableau worksheet;
  7. Press on Size button and shrink your additional pie chart that will serve as a donut hole;
  8. Select one of the axes and mark out a header.

tableau donut chart

Tableau Advanced: Gauge KPI chart



Business people love seeing Key Performance Indexes (KPI) expressed in Donut or Gauge charts. There are no pre-built Gauge chart in Tableau, so users are using their creativity to build it themselves. A Gauge chart in Tableau is usually just a half of the Donut Pie chart.

The sad part of creating a gauge chart using those tutorials is a lack of clarity. Great attention to detail is required to understand the concepts used in those methods. Additionally, you have to ascertain how to implement those methods with your data. Luckily, if you’ll take your time to comprehend Gauge creating concepts, completing your analytic work tasks will be easy.
 

Creation steps for Gauge chart

I am taking Gauge creation concept from Vizartpandey.com blogpost about creating semi donut chart, because it is as simple as possible.

First, we have to understand that default pie chart angle order can’t be used to represent a gauge chart. You can see why in this picture:

Tableau gauge chart angle order

1. Creating supporting data and calculated fields

To change pie chart angle order so it would suit gauge chart, we need to create supporting data that will make new angles.

Sections
Zero_To_Value
Value_To_50
After_50
Remainder
Bottom Half

We add column ‘Number of Records’ and enter ‘1’ for every section row. Also, we do the same with whatever data for which we want to create gauge charts. You can do this addition in Excel easily. This will allow those rows cross-join, when doing FULL-OUTER join on ‘Number of Records’=’1’ in Tableau Data Source window. As a result, we will have 5 different sections for each row of our data-set.

We are going to create a gauge chart that will represent Free Throw percentage of NBA players. Lets use 2018-2019 season stats from https://www.basketball-reference.com.

Now we create calculated fields.

[Max %]– Free Throw Percentage Value based on each player

MAX({FIXED [Player]:MAX([FT%])})

[Arc Angle] – Logical calculations needed to transform our percentage value so that it will be represented in upper half of a circle.

CASE ATTR([Sections])
WHEN “Zero_To_Value” THEN IF ([Max %])<= 0.5 THEN ([Max %])/2 ELSE 0.25 END
WHEN “Value_To_50” THEN IF ([Max %])<= 0.5 THEN (0.5-[Max %])/2 ELSE 0 END
WHEN “After_50” THEN IF ([Max %])> 0.5 THEN (([Max %])-0.5)/2 ELSE 0 END
WHEN “Remainder” THEN IF ([Max %])<= 0.5 THEN 0.25 ELSE (1-([Max %]))/2 END
WHEN “Bottom Half” THEN 0.5
END

 

 

2. Build a Doughnut

To build a Doughnut chart that will be used to create a Gauge chart, we:

  1. Drag ‘Number of Records‘ to Rows field twice;
  2. Set the aggregation of it to minimum;
  3. Right-click on second pill of ‘MIN([Number of Records])’ and select Dual Axis.
  4. In the marks section, select first ‘MIN([Number of Records])’ mark and set the chart type to Pie. For the same mark, drag Sections to Color and Arc Angle to Angle.
  5. Then select second ‘MIN([Number of Records])’ and set the chart type to Circle. Add Max % to label of this circle and drag this label higher so it will be showing in the top part. Change circle color to white and reduce size of the circle. Now we see a doughnut.

3. Finish the Gauge chart

To finish Gauge chart we finish these steps:
  1. Sort Sections manually in this manner (Also assign following colors):
    Sections
    After_50 –Yellow
    Remainder –Grey
    Bottom Half –Any Color (Bottom will be hidden after we’ll fix the axis)
    Zero_To_Value –Yellow
    Value_To_50 –Yellow
  2. Drag Players to Filters card. Press Analysis –> Filters and add Players filter to the right panel. Change Players filter so it will be single value drop-down;
  3. Edit axis so it’s range will be fixed from 1 to 2;
  4. Remove axis headers so we will be left with a Gauge chart.
You can copy numerous of these Gauge charts to the Tableau dashboard. In this case I am duplicating my sheet and moving both Gauges to the Dashboard to compare FT% of two players. Check my Tableau Public dashboard below:

A lovely chart indeed

The pie chart is commonly used in Tableau and for a reason. It’s one of the best Tableau charts you can use. It presents data accurately, shows both proportions and values, and is very easy to interpret. Notably, the classical pie chart is easy to make. Additionally, there are more complicated variations of the Tableau pie chart like Donut chart, Gauge chart and Pie chart on Map. These charts can be used to represent data in BI visualization reports more clearly.

Tableau ,

What is Google Reach around Europe?

2 comments

Almost anyone uses Google or some other search engine nowadays. Those that do not use Google may be very old and poor people. If you have internet, then it’s almost 100% chance that you are using Google, Bing, Yandex, Yahoo, etc. Then is the Google reach number reflect if a country has many old and poor people? In this article, I will analyze Europe countries Google reach numbers against various aspects.

google reach

What is Google reach?

Google describes that reach is an estimate of how many people are in, or interested in, the location you select. It’s based on the number of signed-in users visiting Google sites. This means that Google reach number is how many residents are using google, gmail, google sheets, google disc, etc. in that country on an everyday basis.

In the first two bar charts below I am showing which Europe countries have the best Google reach and reach/population ratio.

Russia and Germany have most people (about 60 million) that use this search engine, despite the fact that Russia’s (140m) population is almost double of what Germany (84m) has. The second chart is representing the reach and population ratio.

 

Why does only 44% of Russian population use Google?

 

Let’s think why do less than half of the Russian people use google sites. Some would say that there are many poor and old people in Russia, so there are fewer people that are using the internet. The average salary in Russia these days is 45100 rubles per month as tradingeconomics.com reports. 45100 rubles equivalent is 700 dollars or 635 euros. It’s not that little in terms of the poorest European Union countries, so it isn’t Russia’s economy’s fault. 

I think that the decisive factor of a bad Russian Google reach and population ratio has to be that many Russians use alternative search engines. Around 43% of people use Yandex.ru search engine in Russia.

Yandex.ru, like Google, provides many services like e-mail, local news, etc., so we can conclude that Russians are content with Yandex quality and don’t need to use Google. This reason has enough weight to exclude Russia from the following analysis.

Below we see TOP8 countries with the biggest reach/population ratio are mostly so called microstates. Usually there are a lot more people living, working and visiting those microstates than actual population. For example, Monaco has a population of 38695, but it has a lot of temporary residents who are not counted for population. Only Netherlands, Cyprus and Croatia are not microstates here, although these countries are popular among tourists.

Looking at the reach/population ratio vs population data, we can see that small population does not mean low reach/population ratio.

reach ration and population

Small Balkan countries (Albania, Armenia, Kosovo) has very low reach rating. Although, we know that those three countries are very poor and very wealthy people lives in Monaco/Luxembourg.

Does wealthy countries have better Google reach / population ratio?

Let’s see how it will play against net income of Europe countries. I am taking median equalized net income data of year 2018 from eurostat and other sources. I drawn the scatter plot in Tableau to see if there are any correlation between measures that I have mentioned (Monaco is excluded because it is a small city state for millionaires).

 

Conclusion: Google reaches richer countries better

From what we see in those two scatter plots, we conclude that there are more google users per population in wealthier countries. R-squared values for trendlines are very convincing for this kind of analysis. R-squared value for Linear regression is 0.396 and for third-degree polynomial regression it is 0.445. Moreover, we could also exclude some borderline countries to make trendline even more accurate. These countries could be: 

  • San Marino (smallest Europe country),
  • Azerbaijan (many people live in poverty and can’t afford internet, there is a big gap between the “middle class” and the poorest people ),
  • Switzerland, Norway, Denmark (old school villagers usually don’t use the internet even though they earn more money)

In conclusion, this is a very interesting finding and I am eager to expand the research. I plan to add countries from the rest of the World to confirm this trend.

SEO, Tableau , ,

Tableau bar chart tutorial

One comment

The fastest way to compare your data is creating Tableau bar chart. In this tutorial, I choose to open Superstore Sales file that I’ve mentioned in my datasets for analysis page. After opening excel file, I choose Orders sheet and drag to data source field.

Selecting measures and dimensions

After connecting with the Super Store Sales file, we see these measures and dimensions: –>

Let’s think what we want to see in our bar chart.

Profit is one of the main KPI’s in business, we are choosing that measure and dragging to rows field. Furthermore, to see what product categories and sub-categories are generating most profit, we need to drag those dimensions to the columns field.

 

tableau bar chart
tableau bar chart

tableau bar chart

Polishing the Tableau Bar Chart

The bar chart is made now, but we have to make it more beautiful and easier to read. Here are the steps that we follow to upgrade our chart:
  1.  Move Office supplies category to the left, because it has most sub-categories. Also move technology category in front of the office supplies, because technology bars are much higher. This way it will look better.tableau bar chart
  2.  Add profit numbers to bars by dragging Profit measure to label field
  3. Now we see that Telephones and Communication category label is not there. Number that would be used for a label is too big, so we decrease text size from 9 to 8.  tableau bar chart
  4. Sort sub-categories by profit (descending).
  5. Add colors to Tableau bar chart to make things further clear. Drag Profit measure to color field. I chose to use red-green diverging for this example. Also, I have marked Use Full Color Range box so that the worst sub-category will be visualized with the richest red in the range. tableau color selection

Final chart view

After finishing our bar chart we can see which categories and sub-categories makes good profit numbers. We see that “the Store” loses money by selling Bookcases and Tables. In a real-world situation, middle managers would be required to report what were the reasons for that. Additionally, as analysts, we can drill-down into subcategories and try to find those reasons in our data.

Tableau ,

Share of interest for different visualization software in USA

One comment

Data was taken from Ubersuggest free keyword tool. We can see from this data that in May 2019 the biggest interest was for Microsoft POWER BI and Tableau software. Although there are many other software that people from USA are interested in. For example, Looker had 100+ k searches, but I excluded it because this keyword can be searched for some different meanings and it is too difficult to estimate what percentage searched exactly for data visualization software called “Looker”.

Tableau , , ,

Tableau dashboard. Portfolio data analysis example

Leave a comment

Tableau dashboard is a place where you can see most important measures ant insights of your data set in one place. Here is an example what I have crawled from my p2p lending investment portfolio data set (about 300+ entries):

As you can see there are 5 different parts from different sheets that tells you something (now my map is lacking a lot of villages and small towns, because I have to enter coordinates for all of it manually). Most of the dashboard is irrelevant for planning investments – it is just a cool presentation of what you might find in data with Tableau.

Tableau , , ,

Tableau blending

3 comments

I will explain how data blending works in this simple Tableau Blending tutorial. I am using fiction data of monthly spending. You can download the excel file here: Steve_And_Karen. The blending operation in Tableau is a left join that can be done on the fly. With data blending you can see instant changes in your visuals when selecting or deselecting possible joins.

Lets say that Steve and Karen are boyfriend and girlfriend, so they are looking at their spending data of last three months to evaluate what their total spending might be. 

Importing data before Tableau blending

In this case you have two data sets (for Steve and Karen) in the different sheets of the same excel file. These are the files for which you want to do Tableau blending. What you have to do after opening your excel file in Tableau is to select on of the excel sheets either Karen or Steve. Then you open your Tableau sheet and press Crtl + D for new data source and select other sheet of the Steve_And_Karen excel file. After that you go to your Tableau Sheet1 and what you have to see is this: tableau blending This means that your sheet is connected to both sides of spending data for girlfriend and boyfriend. Now the first action is to drag Sum, Eu measure to the columns field (remember that we are doing it when Karen part of data is selected as it is shown above). After doing this you only see that Karen’s spending total is 1385 Euros. Now in data selection field you see checkmark at the Karen’s data part icon, this means that visualization is coming from this data source: tableau blending Secondly, we are dragging Karen’s spending and month dimensions to Rows field. Now we can see how she spends money in January, February and March: tableaublending Because we don’t see anything about Steve’s expenditure, we have to select his part of database and drag his Sum, Eu measure to Columns field. Now we see that new right side horizontal bars occurred. The bad thing that we have the same number (1285) for every month and spending type, because it comes from total Steve’s three month expenditures.tableau blending

Blending the data

After dragging Steve’s Sum, Eu to the columns field we see that two grey linking icons appeared: Just press on those crossed links. As you did that, these links became red and visuals in your sheet changed. Now you can see Steve’s side of the chard blended with the primary data of Karen’s. Congrats, you’ve just finished your first data blending. This happened because fields in Karen’s and Steve’s databases had same names, so an option to left join through those data sources (blend) appeared. Below is the final chart (red checkmark indicates that it comes from secondary data source as primary source is Karen’s sheet):
Tableau

Tableau vs Excel: what are the differences?

One comment

Microsoft Excel is one of the most famous and oldest programs for data handling and visualization. On the other hand, Tableau is the new flashy program that more advanced data analysts need to master. There is no right answer to which software is better in a fight: Tableau vs Excel because these two programs have some different features and functionalities. These programs are now bread and butter for pure data analysts.

Tableau vs Excel

Excel vs Tableau: Pros and Cons

Microsoft Excel:

Pros:

  • Program is top-rated;
  • Excel cost is low, most companies and education institutions have it;
  •  Older people are familiar with it;
  •  Wide selection of formulas;
  •  Intuitive manipulation of cells;
  •  You can edit any data in excel easily; 

Cons:

  • Incompatibility with non-Microsoft products;
  • Not designed to work with BIG DATA;
  • Does not have a portal for reports;
  • VBA programming.

Tableau:

Pros:

  • Portal for reports;
  • Interactivity;
  • Modern design;
  • Comprehensive amount of data sources and R programming;
  • Command-line;

Cons:

 

  • High initial cost;
  • You can’t edit data easily like in Excel cells;
  • Complicated data preparation.

Tableau vs Excel: Main Differences

FINDING INSIGHTS

The primary purpose of working with Excel or Tableau is to find critical insights that could not be seen without complex analysis. While working with Excel, you mostly will know what understanding you are searching for from the data you have. However, Tableau lets you blend the data from different sources and browse through a wide selection of insights because it is so intuitive. Because of data blending and drill-down features built-in Tableau, you can spot trends and correlations, and then you can work on discovering what caused them to happen.

VISUALS

When working with Excel, you have to manipulate data on cells level before creating visualizations and presentations. To develop good visuals for Excel insights, you need to be an experienced user.

Tableau lets you create all kinds of good quality visuals with no additional data preparation. Just take some dimensions and measures, drag it to columns and rows, and here you have your recommended visual.

 

AUTOMATION

Most businesses need to live data from multiple sources to make relevant decisions. Excel and Tableau both support live data, but Tableau is better suited for that.

If you want your Excel data to be automatically refreshed, you need to create macros that update all the data when you open your spreadsheet. If you’re going to create or edit macros, you have to know at least basic VBA knowledge. While with Tableau, all the automation processes creation is much more intuitive and less repetitive.

APPLICATION

Tableau is more suited to do business metrics evaluations, while Excel is best for data transformations and new metrics creation. Excel is the best solution while working with relatively small size data, while Tableau can quickly analyze big data problems. While experienced scientists and analysts are still choosing Microsoft Excel for data analysis and visualization, the new trend in business intelligence is to apply Tableau software for finding insights and communicate it to executives of the company.

USE FOR FREE

Tableau, unlike Excel, can be used for free. With Tableau Public You can even use it for free in your company. Just be aware that all the data that you save on Tableau Public is “Public” so be smart and protect any data that can be confidential. Differently, Microsoft Excel is not for free, but most Schools, Universities, Public Libraries, and Offices have this program so you can always find some available version of MS Excel waiting for you.

The solution - Tableau and Excel tandem.

tableau vs excel

What I can say from my experience is that it’s easy to handle datasets in excel before uploading it to Tableau for analysis and visualizations. If you are having a hard time in Tableau not seeing what you are expecting to see, then most probably, you will fix those data source related problems in Excel.
This Tableau and Excel tandem is the best solution for ad hoc analysis that doesn’t have to be updated regularly. If you want daily updates on your Tableau dashboard, it is better to get your data from the SQL server.

Tableau ,