The author first wrote a script to collect and parse all the data. The data were big, so one could not just eyeball them. He did some very basic statistics which a layman like Tuj could discern with ease. After seeing some patterns and forming some opinions, he decided to take a deeper look. At this point more tools from statistics were needed and at this point things became not so transparent for Tuj who had not undergone serious training in statistics. However Tuj knew that he could understand if he read about the theorems on which the tools were based. He believed that it mostly boils down to linear algebra and the invariants are attached meaning for the data. He took note of the names and added those really black boxes to his tool set. He will surely forget when the need arises. In practice, most often it is not necessary to understand the principle behind the tools. People solve problems fast by having a large pool of tools to choose from and knowing which are the right tools for the problems. This is a feat in itself. However knowing why the tools work make it possible for one to hack the tools and to apply them in novel ways to tackle what one could not solve. Tuj also thought about that so much of what he knew was completely unrelated. He felt light-years and light-years removed. However such distance melts like butter, unlike those intangible barriers that have been barring him from entering a neighbouring domain. Back on topic, let us take note that the author still needed to choose efficient algorithms to carry out the analysis.
These days Tuj talked to people and it seemed that the same words carried different meanings for us so that we had to keep clarifying ourselves in order for the conversation to go on, extremely slowly on... Tuj got confused on parts unanticipated by people and their explanation was not to the point because they did not know what confused Tuj! Then after much toil, we managed to click. The foreign language finally connected with Tuj's system. What satisfying moment!