Open waredjeb opened 9 months ago
Are we happy with an output like this? @rsreds @felicepantaleo @bucket420 Or do we want something more sophisticated?
# wind angle1 angle2 height fitness0 fitness1
0.177660754571031643,0.224173540480451322,0.418394021075605638,115.050193126576274949,0.099999999999999978,0.000000000000000000
0.203423985357212678,0.260409003579986753,0.396187104120431510,114.250420772215420584,0.099999999999999978,0.000000000000000000
I'd stick with csv style for the header as well. With or without the #. You still have to treat it carefully when reading it if you have the header.
Ok, so probably I'll produce the following, keeping the same separator for values and header
wind,angle1,angle2,height,fitness0,fitness1
0.177660754571031643,0.224173540480451322,0.418394021075605638,115.050193126576274949,0.099999999999999978,0.000000000000000000
0.203423985357212678,0.260409003579986753,0.396187104120431510,114.250420772215420584,0.099999999999999978,0.000000000000000000
Can be easily read with pandas:
d = pd.read_csv("pareto_front.csv", header=0, sep = ",")
It would be nice to also dump the name of the parameters together with their values to the final
csv
file (mostly for CMSSW configuration)For CMSSW: maybe it would be also nice to save the final CMSSW configuration dump with the updated parameters
@rsreds you can assign me the issue if you want, I wanted to do that already