Does the 'line-grouped model' perform better? about etrago HOT 5 CLOSED

Here are the times for the SH Status Quo scenario with 1000 snapshots (2420-3420):

Non grouped:
--> Import: 0.092
--> Calculations: 24.004

Grouped:
--> Import: 0.118
--> Calculations: 17.601

from etrago.

Got one more, this time with SH Status Quo and 3000 snapshots (crashed twice, because I ran out of RAM):

Non grouped:
--> Import: 0.199
--> Calculations: 123.333

Grouped:
--> Import: 0.227
--> Calculations: 79.46

from etrago.

The line grouping does indeed change the runtime of the calculations. I've done 3 runs with 2 calculations each (non grouped & grouped). The "Import section" included the line grouping process:

First run:
Non grouped:
--> Import: 0.695 min
--> Calculations: 39.955 min

Grouped:
--> Import: 3.954 min
--> Calculations: 31.911 min

Second run:
Non grouped:
--> Import: 0.686 min
--> Calculations: 42.436 min

Grouped:
--> Import: 4.336 min
--> Calculations: 37.944 min

Third run:
Non grouped:
--> Import: 0.566 min
--> Calculations: 32.261 min

Grouped:
--> Import: 2.942 min
--> Calculations: 25.872 min

Even though the calculations are fairly quicker, the grouping of the lines seems to eat up the time gained. Summed up the difference seems to be fairly small...

from etrago.

Could you do some testing with more than 2 snapshots? May be with the sh scenario (faster)
I would expect that the calculation time will then outrun the extra time at the beginning?!

from etrago.

Each calculation was done with 20 snapshots. I can try the SH scenario with 500 snapshots tomorrow.

from etrago.