{"id":3596,"date":"2021-06-22T20:33:38","date_gmt":"2021-06-22T20:33:38","guid":{"rendered":"https:\/\/cloudxlab.com\/blog\/?p=3596"},"modified":"2021-06-22T20:36:04","modified_gmt":"2021-06-22T20:36:04","slug":"parallel-computing-with-dask","status":"publish","type":"post","link":"https:\/\/cloudxlab.com\/blog\/parallel-computing-with-dask\/","title":{"rendered":"Parallel Computing with Dask"},"content":{"rendered":"\n
\"Dask
Source: dask.org<\/figcaption><\/figure>\n\n\n\n

I recently discovered a nice simple library called Dask. <\/p>\n\n\n\n

Parallel computing basically means performing multiple tasks in parallel – it could be on the same machine or on multiple machines. When it is on multiple machines, it is called distributed computing.<\/p>\n\n\n\n

There are various libraries that support parallel computing such as Apache Spark, Tensorflow. A common characteristic you would find in most parallel computing libraries you would is the computational graph. A computational graph is essentially a directed acyclic graph or dependency graph. <\/p>\n\n\n\n\n\n\n\n

Say, for example, we want to compute the value of Z from x and y and we are giving this:<\/p>\n\n\n\n

y1 = y*4<\/p>\n\n\n\n

x = y1*y1 + 3<\/p>\n\n\n\n

Z = x\/3 – 4y1<\/p>\n\n\n\n

This computation can be expressed as a dependency graph:<\/p>\n\n\n\n

\"\"
Directed Acyclic Graph representing computation<\/figcaption><\/figure>\n\n\n\n

Z depends on x and y1 and x depends on y1 and y1 depends on y. This graph is evaluated only when you need not automatically. This helps in optimization.<\/p>\n\n\n\n

Here is an analogy of lazy evaluation (Starts at 1 min.):<\/p>\n\n\n\n

\n