Codetown ::: a software developer's community

ResourcesLast week, we went over higher order functions in Kotlin. We learned how higher order functions can accept functions as parameters and are also able to return functions. This week, we will take a look at lambdas. Lambdas are another type of function and they are very popular in the functional programming world.
Computer programs are made up of two parts: logic and data. Usually, logic is described in functions and data is passed to those functions. The functions do things with the data, and return a result. When we write a function we would typically create a named function. As we saw last week, this is a typical named function:
fun hello(name: String): String {
return "Hello, $name"
}
Then you can call this function:
fun main() {
println(hello("Matt"))
}
Which gives us the result:
Hello, Matt
Functions as DataThere is a concept in the functional programming world where functions are treated as data. Lambdas (functions as data) can do the same thing as named functions, but with lambdas, the content of a given function can be passed directly into other functions. A lambda can also be assigned to a variable as though it were just a value.
Lambda SyntaxLambdas are similar to named functions but lambdas do not have a name and the lambda syntax looks a little different. Whereas a function in Kotlin would look like this:
fun hello() {
return "Hello World"
}
The lambda expression would look like this:
{ "Hello World" }
Here is an example with a parameter:
fun(name: String) {
return "Hello, ${name}"
}
The lambda version:
{ name: String -> "Hello, $name" }
You can call the lambda by passing the parameter to it in parentheses after the last curly brace:
{ name: String -> "Hello, $name" }("Matt")
It’s also possible to assign a lambda to a variable:
val hello = { name: String -> "Hello, $name" }
You can then call the variable the lambda has been assigned to, just as if it was a named function:
hello("Matt")
Lambdas provide us with a convenient way to pass logic into other functions without having to define that logic in a named function. This is very useful when processing lists or arrays of data. We’ll take a look at processing lists with lambdas in the next post!
Tags:
Codetown is a social network. It's got blogs, forums, groups, personal pages and more! You might think of Codetown as a funky camper van with lots of compartments for your stuff and a great multimedia system, too! Best of all, Codetown has room for all of your friends.
Created by Michael Levin Dec 18, 2008 at 6:56pm. Last updated by Michael Levin May 4, 2018.
Check out the Codetown Jobs group.
In this podcast, Shane Hastie, Lead Editor for Culture & Methods, spoke to Satish Kothapalli about the transformative impact of AI and vibe coding in life sciences software development, the acceleration of drug development timelines, and the evolving roles of developers in an AI-augmented environment.
By Satish Kothapalli
Unlock the power of AWS Lambda Managed Instances, seamlessly combining serverless functions with Amazon EC2 for optimal performance and cost efficiency. Designed for steady-state workloads, this solution automates instance management, reduces cold starts, and enables multi-concurrency.
By Steef-Jan Wiggers
Louis Ryan shares a compelling vision for modern cloud native hybrid networking. He critiques primitive network abstractions (the "Big IP" problem) and rigid security policies that rot and cause SPOFs. Discover how architects can elevate network functionality, bake in identity (mTLS/PKI), and leverage composability to achieve repeatable policy enforcement everywhere their applications run.
By Louis Ryan
Grab updated its internal platform to monitor Apache Kafka data quality in real time. The system uses FlinkSQL and an LLM to detect syntactic and semantic errors. It currently tracks 100+ topics, preventing invalid data from reaching downstream users. This proactive strategy aligns with industry trends to treat data streams as reliable products.
By Patrick Farry
Serving Large Language Models (LLMs) at scale is complex. Modern LLMs now exceed the memory and compute capacity of a single GPU or even a single multi-GPU node. As a result, inference workloads for 70B+, 120B+ parameter models, or pipelines with large context windows, require multi-node, distributed GPU deployments.
By Claudio Masolo
© 2025 Created by Michael Levin.
Powered by