I would add it is because of the popularity of the packages that are used in data science. Comparing the numbers of modules in all package managers Python doesn't stand out. First there is NPM for Javascript way above and then there is everyone else.
I think there is just too much ceremony involved in OO languages to get basic math concepts working. You don't need to know about polymorphism, inheritance, abstract classes, constructors, interfaces static & instance data types...if all you want to do is crunch some numbers.
I enjoyed learning Python a lot more than I enjoyed learning subtle differences between C# and Java.
Some other stuff I really like in Python:
Parsing JSON. It feels so natural. In Java & C# I would spent 1 day to find the best framework, then create all the objects to map JSON data, in Python JSON is just a dictionary data structure. It is 1 line to convert JSON string into a dictionary and then you just use the dictionary syntax to read/write data.
It teaches you to write tests. You need tests for everything, knowing that there is no compiler to back you up, your code coverage needs to be a lot higher than when working with strongly typed languages.
You just open a file and start writing code, without the need to declare any classes, static main functions, configure the project with what to run etc...
Agree. But we are not trading firms building platforms for thousands of consumers. So to each his own
Sorry for hijacking the thread @TheGuru12