Mappers and Reducers.
Every computation can be expressed as a combination of mappers and reducers, and Big Data™️ leveraged this in a big way, because if you can express everything as mappers and reducers, you can parallelize on each mapper and reducer.
The functional programming community further shows that every mapping function is a reducer.
map f = reduce (concat . f) []
All computations are reducers.
Therefore: your 1 instruction set computers (OISC)
And if you only need 1 instruction, you need 0 instructions (ZISC)
with the following formula
map ((1),(k "Cryptocurrency"),(3),(4))
SIMPLE! We don't even need to write a program to do this transformation, as simple spreadsheet formula handle this directly.
But how do we then reduce the spreadsheet of multiple crypto across multiple blockchains to 'just the crypto, please'-spreadsheet, something like this:
Let's review three approaches to this desired output.
- One approach is to work within the spreadsheet, itself, using its functions, such as sumif() or dsum().
The problem I have with this approach is that you're using the spreadsheet as a database, and you have to learn these functions, which is learning a new language, really.
- Next, you can write a problem in, e.g. Rust, to take on the mappings and reductions.
I'm digging a modular approach here, where you code individual mappers and reducers, and you can plug-n-parallelize to your heart's content.
- As I've already codified my Blockaverse portfolio as an ontology, I simply query the ontology and receive a graph as a result.
Dude. 😎
Cryptocurrency pop-quiz
Choose one of the approaches above, or, forge your own path and create your own, then solve the problem of mapping a cryptocurrency portofolio and reducing it to tokens and their USD amounts across the Blockaverse.
DOIT! TOIT!
- (answer)
No comments:
Post a Comment