₦airaland Forum

Yo fam!!! I'm back again with another form of visualization; this time Power BI. Few days ago I did some analysis on the "Students Levels online" dataset using my SQL and also created an interactive Excel Dashboard which I posted (if you haven't seen it, please check it out)

So today I will sharing my analysis with this data using Power BI. I love using Power BI for my visualization. It gives me more and better options than Excel. With Excel you need a slicer to analyse the data to get more insights but with Power BI, the data can be filtered without the slicers. Clicking on a particular value on the dashboard can get you what you want.

Without wasting much of your wonderful time. Here We Go........

A) The first pics shows the complete dashboard without any filtering. This is everything you need to know about this data. With the help of the data labels you can also read and understand what is happening on the dashboard.

More and More Filtering

e) The last two pics showed just one selected item from one slicer but with the pics below more selections from different slicers was done to give you more insights from the dashboard. This pics shows the number of students from "RURAL" location that said "YES" to doing their IT (industrial training)

F) And finally with more selections, this pics shows the the number of students that used the "MOBILE" device, with internet type "MOBILE DATE", which has a "4G" network type to get access to their online education.

And apart from just the total number of students or total number of male and female students, other analysis like the age distribution, the number of students from each education level, the number of students from each institution type can be gathered from the dashboard etc.....

The slicers helps you analyze the data how you want. And looking at the dashboard, a lot more analysis and insights can be gathered.

#Dataanalyst
#Dataanalysis
#Datavisualization

Peace.

MORE FILTERINGS
c) The third pics shows the "High" flexibility levels of the students

d) The fourth shows the Network types used by the students. The selected item for this analysis was the "4G" network. From the dashboard we are able to figure the number of male and female that used 4G, the institution and education level that used 4G etc...

Few days ago i shared with you some of the analysis i did on a dataset which i got from kaggle, using MySQL. A lot of analysis was done on that "student levels online" dataset; if you haven't seen it plz look at my last topic or thread.

For most technical users, MySQL is enough to get insights from a data but for some, not really. A graphical representation also known as DASHBOARD was needed to convey the insights gotten from the data to both technical and non-technical users, which in turn helps the users make more informed decisions.

Please Note: This is an INTERACTIVE DASHBOARD.

a) The first pics shows the full dashboard without any filtering. Not much can be gotten from this dashboard coz it mostly shows the Total of some of the values from the dataset. For more in-depth analysis we...…..

b) FILTER. Yes we filter. Filtering the data gives you the freedom to explore more options. You are able to analyze the data based on the particular insights that you need. For example the second pics shows the financial condition of rich students. And as you filter a particular item from the slicer, the data changes. The more you filter the more the data changes, giving you correct analysis of what you are looking for.

Note: The slicers were modified. The dark yellow shows unselected item with data and the purple shows unselected item with no data. The selected item was the "rich" value.

airsaylongcome:
Have you seen this?

careers dot hotjar dot com/o/sales-data-analyst-emea

€55,000 to €75,000 annually

No. I am gonna apply ASAP. Thanks Man.

airsaylongcome:
Nice stuff. I like how you regularly throw these kinds of analysis out here. I'm more shocked at the level.of detail that "they" can tell what kind of connection (4G) the devices are on. Meaning they can tell to very huh detail what device is been used, where in the world it is been used from. Bye bye privacy.

Personally, I prefer these kind of quantitative analysis to Dashboards o. I can see the raw figures and make my own "back of the napkin" calculations.

Are you aggressively hunting Data Analysts roles? I think you should be if you aren’t


Also do something to aggregate this into the various types of connections, without disaggregating it into Uni, School and College.

Dataset looks like UK dataset

Boss! thanks for the input. Like i said earlier a lot more analysis was done with and without aggregating them into uni, school and college or the institution type. The screenshot is a lot, 40 of them or so. I wish I could post them all. And yes I am aggressively hunting for data analyst roles both naija and abroad.

That is all I can share. More analysis was done on the location, flexibility level and financial condition of this students. I feel like I have posted enough. I am going to stop here. This is why a lot prefer dashboards to all this.

Note: This is a simple analysis, nothing advanced. The "group by clause", "order by clause", "count" and "concatenation" was used for most or all of the analysis.

Next is to design a simple DASHBOARD.

Peace.

5) LETS LOOK AT THE INTERNET AND NETWORK TYPE USED BY THIS STUDENTS

a) The first pics shows the internet type used by students from each education level. And from the data below we can see that the WIFI is the most internet typed used by university and college students. While school students use mobile data more than the WIFI. Also university students use more WIFI than any other education level and school student use more of mobile data.

b) The second shows the network type used by students from each educational level. 4G is the most used network type. Of course it is....

4) This analysis show the number of students that attended school, university and college from each institution (private and public). And from the result below we can see that there are more school and university students from private institution. While college has more students from the public institution than the private.

3) NOW LETS LOOK AT THE STUDENTS THAT WENT FOR THEIR IT (industrial training) AND THE ONES THAT DIDNT.

a) The first pics shows the the total number of student that said YES to doing their IT and the ones that said NO. Done with their education online, majority of students had refused to go for their IT. Would this have happened if the classes were offline?

b) The second show the students based on each education level that said YES or NO to doing their IT. This analysis show that unlike students from the school and college level, more university students choose to go for their IT. Do you know?

c) The third shows the students based on each institution that said YES or NO to doing their IT. And as you can see both institution has the highest number of Students saying NO to any IT. But more students from private institution said Yes to IT than the students from public institution.

2) NOW I WANT TO KNOW THE DEVICES USED BY THIS STUDENTS FOR THEIR ONLINE EDUCATION.

a) The first pics shows the devices and the total number of students that uses each devices for their online classes. And as you can see, the mobile is the most used device. Tech is advancing. students feel more comfortable using their mobile for classes than the computer or tab.

b) The second shows the devices used by students from each educational level. And as you look closely you would see that the mobile device is the most used device on all levels. Also the school happens to have the highest number of students using the mobile device.

c) The third show the devices used by students from each institution. And just like the second pics, the mobile device is the most used device from both institution. Also the private institution has more students using the mobile device than the public.

LET THE ANALYSIS BEGIN.....

1)
a) The first pics shows the total number of students from each educational level (university, school and college). The summary column is the concatenation of the first two columns to make it look more explainable and presentable to the anyone looking at the data.

b) The second pics shows the break down of both male and female students from each educational level. And this analysis shows you that there are more male students from the university and school and less at the college i.e. more female students than the male student

c) The third shows the number of students from each institution. And as you can see there are more students doing their online education at a private institution than the public institution.

d) The fourth is also a break down of both male and female students doing online education at the private or public institution. And from the analysis there are more male students at both the private and public institution. The boys are really taking the online education more better than the female.

This is a "student adaptability levels" dataset which has been on my system for a while. Feeling enthusiastic and hungry for data to analyze I decided to work on this dataset.

I got this dataset from kaggle, loaded it on excel for data cleansing but was quite disappointed coz there was nothing to clean. So I imported this already clean dataset to MySQL for more analysis.

This is what the dataset looks like...…

Keskyramirez:
How do I learn data visualization and what basic skill is required to learn it

YouTube, medium and Ebooks Bro. Lots of good stuff there. That is where I learnt mine before advancing to online courses (mostly for the certifications). Just type Data visualization and the lessons that you need shall be added unto you.

Sageez:
Any idea where I can volunteer?

No man. You should try freelancing on upwork. The more offers you get, the more skills you learn, which should get you ready for a job.

Sageez:
Aside Excel, which other software are you good at?

For me, Excel, sql, python and R, but I still feel I am not ready yet. I want to get myself grounded then intern or volunteer for any firm before applying for jobs

Excel and sql is where my strengths are. Python and Power BI; i'm goood but not like the former.

You should volunteer. No one is ever ready. You just have start and keep learning on the job.

NotBeenPaid:

---

Well done.

---

Thank you.

Sageez:
I have finally gone through all your chart but sorry, I am not impressed.
I am not saying this to hurt you but to make you work more and harder.
Please study graphs and chart and understand the reason to use them.

Let's take bitcoin for instance, the trend changes almost every minute, the best chart to display this will be line chart.
But let's say we want to know the performance of boys against girls in a particular subject, the best for this should be a histogram or bar chart.

I like your courage. We won't stop learning till we blow.
Sorry if you fined my first sentence offending.

Lol. Not at all. We are all here to learn. Thanks for your input.

Sageez:
Sir,

For the first dashboard, a histogram or bar chart would have been better.
Line charts are used for showing trend, but for categorical variables, histogram is usually the best.

Please correct me if I am wrong cuz I am also here to learn

Yes, you are absolutely correct. There are rules to data visualisation. Every chart has its purpose and must be used the right way. But on this dashboard, the emphasis was on the data label below the chart and not the chart itself. When creating this dashboard I'd imagine the stakeholder or client asking for a more labeled dashboard and Based on personal experience I have worked with clients who are only interested in labels not the chart itself. And that is why I opted to using Scatterplot on all boards and focus more on giving a well detailed label.

This is where I press the stop button There were more results I got from analyzing this dashboard but I can't share them all. I think the ones above are enough

Thank you for your time. Peace.

#dataanalytics
#dataanalyst
#datavisualization

6) And the just like the one above, I was also able to get the airline, number of flights, hours and price of flights that made Zero stops using the STOPS slicer.

5) Using the STOPS slicer this dashboard shows the airline, number of flights, cities, hours and price of flights that made two or more stops before getting to their final destination.

4) This is just like the dashboard above but with a different selection. This time I filtered the data to show me the airline and flights that left in the morning and arrived at night.

3) From this dashboard I was able to filter the data to show me the airline and flights that left early in the morning and arrive arrive early in the morning. This also shows the source cities and destination cities of flights that left early in the morning and arrive early in the morning. You also see the price made from this kind of flight, and also the number of hours spent on air.

Note: I also modified the slicer to make it more explanatory and attractive
* clicking on a selected item with data shows an Ash color
* the unselected item with data shows Blue.
* and finally the unselected item with no data shows Yellow.

2) The complete Dashboard without any Filters (i.e. using the slicer)

For the past few weeks I have been visualizing my data using Python and Power BI coz that has been the requirements from most client. I had forgotten you could also visualize your data using Microsoft Excel .

So today I got on Kaggle, got some sample data and worked on it.

This is a flight dataset set that shows the flight, source city, departure etc..... of each Airline.

The purpose of this analysis is to know the Airline(s) and cities that are dominating the aviation industry in India and how much revenue is been created from this airlines and cities.

So Here We Go....

1) The Flight Dataset

5) This pics shows Subquery using the where clause. The result of this query gives you genres and publishers not in the common genres and publishers from both tables. Hope you understand. Lol.

This is me just messing around with joins and subqueries.

4) This pics below shows Subquery using the From clause. The first pics (result) of this query gives you the not common genres and publishers from both tables and the second pics shows the common genres and publishers from both tables.

3) More filtering like the picture above but with # in it. The # in front of a query means that line won't be executed. Instead of writing queries over and over again, using the # kinda helps and saves times. All I need to do is put # in font of the queries I Don't want executed.

2) The pictures below shows filtering of the joined tables using the where, Group by and order by clause.

One of the queries I try to avoid most times in SQL is the subquery. Not the simple ones, the ADVANCED Subquery. Lol. Joins are quite easy to work on.

A subquery is a query that is nested inside a SELECT , INSERT , UPDATE , or DELETE statement, or inside another subquery.

When I am not working on a real dataset for clients, I do go on kaggle to get open dataset to work on.

I am gonna be sharing with you joins and subqueries I just worked on. It's ADVANCED but not deep like that. Lol.

This a Game dataset with two different tables.
(I) ps4_gamessales table
(II) xboxone_gamesales

The picture below shows how I created the database "Game_sales" and the joining of both tables and the result.