Computing your first FK tables

Once PINEKO is correctly configured (for a detailed guide on how to set up pineko see here (opens in a new tab)), one can start computing FK tables. This requires three main steps:

1. Write the yaml file for the wanted dataset and check if all the needed grids are in correct folder.
2. Compute the operator cards and the related eko operators.
3. Compute the FK table.

1. Yaml file.

The first step depends a lot on what kind of dataset we want to construct. For a simple dataset, like the one shown in the Inputs section of here (opens in a new tab)), it is enough to write

conversion_factor: 1.0
operands:
  - - HERA_NC_225GEV_EP_SIGMARED
operation: 'null'
target_dataset: HERACOMBNCEP460

where the name of the dataset, that will be used from now on, is HERACOMBNCEP460. Notice that the name of the yaml file needs to be the same of the dataset (so in this case HERACOMBNCEP460.yaml). In this case we are using only one grid (under the section operands) and no operations are performed on it (an example of a possible operation is ratio but of course it requires two grids at least). Once we are sure that the grid HERA_NC_225GEV_EP_SIGMARED.pineappl.lz4 is in the correct folder, as provided by the pineko.toml file, we are ready to proceed to the next step.

2. Operator card and eko.

In order to compute the operator card, it is enough to run¹:

pineko theory opcards <theory_number> <dataset_name>

where <theory_number> is the number of the theory card that we want to use (and that has been previously configured as explained here (opens in a new tab)) and <dataset_name> is just the name of the dataset for which we want to compute the operator cards (for our example it would be HERACOMBNCEP460). An example of the output of the command

pineko theory opcards 208 HERACOMBNCEP460

is:

Analyze HERACOMBNCEP460
Success: Wrote card with 23 Q2 points to
data/operator_cards/208/HERA_NC_225GEV_EP_SIGMARED.yaml

where the <theory_number> is 208 in this case. Note that if the yaml file contained more than a single grid, the command would iterate over all the grids, producing an operator card for each of them. This is because indeed we need an eko operator for each grid and, at the end, a FK table for each eko. Having written the operator cards, it is time to compute the ekos. To do so, it is enough to do:

pineko theory ekos <theory_number> <dataset_name>

where <theory_numbers> and <dataset_name> have the same meaning of before. This most likely will take some time (or a lot of time): play some games in the meanwhile (or work on something else if you prefer). The EKO output is pretty long (and you can read it all from the logs) but, if everything just worked fine, you should see something similar to this line at the end of the log file

2022-10-18 10:22:09,873 pineko.theory/INFO: Finished computation of HERA_NC_225GEV_EP_SIGMARED - took 1127.056456 s

Good job! Now you have everything to finally compute the FK table.

3. FK tables

Even for computing FK tables the PINEKO command is pretty simple

pineko theory fks <theory_number> <dataset_name>

and, even in this case, this could take a while. The output of this command should be something similar to

2022-10-18 10:32:14,240 pineko.theory/INFO: Start computation of HERA_NC_225GEV_EP_SIGMARED
2022-10-18 10:32:14,241 pineko.theory/INFO: max_as=3, max_al=0, xir=1.000000, xif=1.000000
2022-10-18 10:32:16,500 pineko.theory/INFO: Finished computation of HERA_NC_225GEV_EP_SIGMARED - took 2.259366 s

Notes