Graph Lab implements the MapReduce Abstraction through an Update Method which is the "per-vertex update" and a sync function which is the "global reduction". The Update function is able to read and modify overlapping sets of data/program state in a controlled fashion through "powerful scheduling primitives". The GraphLab site is here: http://www.select.cs.cmu.edu/code/graphlab/
This comment was marked helpful 0 times.
jinghuan
What does lazy execution mean??
This comment was marked helpful 0 times.
nrchu
I believe lazy execution is waiting to evaluate an expression until the value is needed, and also tries to do sharing of values for repeated expressions as much as possible.
Graph Lab implements the MapReduce Abstraction through an Update Method which is the "per-vertex update" and a sync function which is the "global reduction". The Update function is able to read and modify overlapping sets of data/program state in a controlled fashion through "powerful scheduling primitives". The GraphLab site is here: http://www.select.cs.cmu.edu/code/graphlab/
This comment was marked helpful 0 times.
What does lazy execution mean??
This comment was marked helpful 0 times.
I believe lazy execution is waiting to evaluate an expression until the value is needed, and also tries to do sharing of values for repeated expressions as much as possible.
This comment was marked helpful 0 times.