Exploring interests of developers with analysis and comparison of the trends with Indian developers

Hi!

I’m Vivank Sharma, an undergrad student of Data Scientist Nanodegree of Udacity.

This article is about finding from my project of ongoing nano degree where I have analyzed the developer’s feedback and drawn up answers to some questions about Indian developers.

I have used Stack Overflow’s annual Developer Survey. It’s the largest and most comprehensive survey of people who code around the World. Each year, they field a survey covering everything from developers’ favorite technologies to their job preferences. In the year 2019, nearly 90,000 developers took the 20-minute survey, which you can find here.

With the approach of new territories in India, for example, Artificial Intelligence and Data Science, many programming dialects that were inadequately spoken and utilized are currently encountering an ascent in fame.

You may have just heard Python is the programming language existing apart from everything else Functional Programming is far superior to Object-Oriented.

In this article, we will examine factual information to check if these programming dialects are genuinely being utilized in India or the event that they are simply bits of gossip.

Why the analysis?

I am being an aspiring developer that has always been one question wondering in my mind about what to learn next?

The answer to this question depends on various factors like:

• What type of developers is currently most in the World?
• What kind of developers is now most in our own country?
• What type of developers is getting paid the most?
• Which programming language is used most in the Industry?
• Which programming language current industry wants to learn?

For this, we are going to utilize information from Stackoverflow’s 2018 and 2019 Annual Developer Survey.

Consistently, Stack Overflow directs a gigantic review of individuals on the site, covering a wide range of data like programming dialects, employments, code style, and different other data.

In this immense ocean of the developer community, we always tend to wonder about one question specifically :

What type of Developers is most in the whole World?

For gaining this information, we first had to pick out the DevType column from the dataset.

From the above output, we can see that an individual has chosen multiple options as DevType.

Hence, we need to extract information by using separator ‘;.’

For that, we will use a split column function, as shown below:

Once we have our cleaned data now, and we can visualize the data and observe it.

For visualizing, we will use the plot function, as shown below:

We got the following findings after analyzing the dataset for DevType.

• Global DevType 2018:
  

• Global DevType 2019
  

We can observe that from the above graph that in 2018 the most type of developers was Back-end Developers. Still, in 2019 we can see a drastic growth of Full-stack Developers, which shows that surge in demand for various skills and not limited to one ability, which increases the dependency.

Also, we can note one more thing about the drastic surge in the developers of desktop or enterprise applications in the year 2019.

Now, we will see the trend of DevType of India.

For that, we will first need to filter out the dataset using the country.
Which can be done by our filter country function, as shown below:

We got the following findings after analyzing the filtered dataset for DevType.

• India DevType 2018:
  

• India DevType 2019:
  

We can observe that the developer type ratio of India compared to the global Developer type ratio isn’t much different. However, we can see that the number of Full-stack developers was still behind the Back-end developer but not by very much and would have surpassed them by now.

Well, we have till now concluded that the trend in DevType of India isn’t much different than that of Global.

Now, one of the most important questions which need to be answered is:

Which type of Developer earns the most according to Stackoverflow’s survey?

Now analyzing that is a tricky part as we have to split the DevType and then remap the salary of an individual to those DevType correctly.
So, at first for splitting the DevType with indexing to each individual, we will use split_column_value function as shown below:

and after that, adding the correct salary according to the individual can be achieved by our add_salary_column service as shown below:

As our data is ready for plotting now we will use groupby and plot function of pandas to plot the histogram like mentioned below:

We got the following out from our analysis:

• Highest Paid 2018
• Highest Paid 2019

Here, we can observe that if we leave the top 3 management positions, we can see that the most paid developer skills are among DevOps specialist, Data Scientist or machine learning specialist, or data engineer.

The result of the year 2018 and 2019 is almost the same except the Business Analyst role, which signifies that companies are starting to adopt data science to practically every aspect, and there has been a growth of advanced data science techniques like machine learning and deep learning.

Any analysis isn’t complete without knowing:

What’s the preferred language of our Developer’s community is and what they want to learn in the future?

The analysis of the preferred language is quite easy and can be done using two functions, as shown below:

The year 2018:

• Top 10 programming language used by the developer in 2018
• Top ten programming language which developer desire to learn in 2019

The year 2019:

• Top 10 programming language used by the developer in 2019
• Top ten programming language which developer desire to learn in 2020

We can see the hype of the top 10 programming languages in the graph above for the year 208 and 2019.

Though a notable difference that we can see is Python is more famous than Java in the year 2019.

We can observe that in 2018 people wanted to learn JavaScript and Python, which resulted in the growth of Python in the year 2019.

In the year 2019, the top 5 scheme and desire remains the same.

But on a note, we can see the increase in the willingness to learn ‘Go’ is also increasing, which can suggest a future era of ‘Go’ language.

You can download the whole repository from here.