Alhamdulillaah I finally completed section 1.2 of the SICP classic. I was amazed I took almost 150 days to complete 61 pages. I definitely need to sit up! Reading the text was not challenging, the bottleneck was ~~‘forcing’~~ getting myself to finish the exercises.

I was going to start on section 1.3 but I felt it would be better to reflect on the knowledge gained first.

**1. A deeper understanding of procedures and their processes**

Not all procedures are created equal – seemingly recursive procedure can be iterative in execution. Moreover, there is nearly always a way to make a recursive procedure execute in an iterative manner. This can lead to speed gains and also help you handle HUGE data.

Programming is deeper than I thought…

**2. Mathematics**

My Mathematical forays are usually related to Computer Science or Machine learning however I got to dabble into new fields. First was the Ackermann function; a fascinating recursive function which grows with mind-boggling speed – you have to think hard to grasp this. Next came primality testing and probabilistic approaches to prime number verification.

The Fermat test is good enough however it is fooled by the Carmichael numbers. To be sure, use the Miller-Rabin test.

I do not know how much of these will be useful in real-life but yeah… it is good to know.

**3. Algorithms – exponentiation**

The rapid exponentiation algorithm was an eye-opener – do stuff twice as fast :). Once you implement the ‘speed-up’ and ‘slow-down’ functions and handle all cases properly, you can take down exponentiation from a *O(n)* operation to a *O(log n)* operation. Combining this with a ‘tweaked’ recursive-but-iterative-in-execution procedure leads to ‘tales’ of joy…

It was also interesting to see how minor changes to code could wipe out performance gains made from clever algorithms. Exercise 1.26 showed how easy it was to lose the *O(log n)* gains from exponentiation by doing irrelevant work. A subtle ‘refactoring’ might have huge implications…

**5. Perseverance**

I initially find most exercises daunting and struggle to understand the task. I *force* myself to think about the problem for at most 15 minutes; if I still do not get it, then I allow myself to look at available solutions.

Alhamdulillaah I usually figure out the solution during the time window and then look up existing solutions to see other problem-solving approaches. I have also looked up solutions when I got stuck too – I was seeking ‘*inspiration*‘ :). It’s great to know we can solve most problems if we only persevere insha Allaah.

And that’s about it! Section 1.3 has about 18 exercises; since I typically solve an exercise in about 2 – 3 days (I have a 25 minute daily study schedule), I hope to be done with this section in about 4 – 6 weeks insha Allaah. Watch out for a new update then insha Allaah.

Here are my solutions on Github.

Here are a couple of my more academic-themed musings

3. MOOC Review: Machine Learning