Create lines more general approach including 3ph params

[elodin]

This alters the behavior of create_lines() based on the discussion on #1382.

The function will now use whatever parameters are specified in the std_type(s) instead of only using those parameters directly mentioned. It works with either multiple lines and one std_type as well as multiple lines and multiple std_types. Also the bus inputs can now be int, for convenience.

Codecov is showing an error, because the function is shorter regarding the number of code-lines now, which changes the ratio of covered to uncovered lines of the file. All lines of the function are covered by tests, though.

Tests of the changes have been conducted using the following code:

---------- CODE ----------

import pandapower as pp
import pandapower.networks

#Import test network
net = pandapower.networks.mv_oberrhein("generation")

#zero seq. parameters added to only one specific std_type
net.std_types['line']['NA2XS2Y 1x185 RM/25 12/20 kV']['r0_ohm_per_km'] = 123
net.std_types['line']['NA2XS2Y 1x185 RM/25 12/20 kV']['x0_ohm_per_km'] = 321
net.std_types['line']['NA2XS2Y 1x185 RM/25 12/20 kV']['c0_nf_per_km'] = 456
net.std_types['line']['NA2XS2Y 1x185 RM/25 12/20 kV']['g0_us_per_km'] = 654

#Line copied and droped
line = net.line.loc[0]
net.line = net.line.drop(0)

#add one line based on formerly droped line
pp.create_lines(net,
from_buses=line['from_bus'],
to_buses=line['to_bus'],
length_km=line['length_km'],
std_type=line['std_type'],
name=line['name'],
df=line['df'],
parallel=line['parallel'],
in_service=line['in_service'])

#more lines copied and droped
line1 = net.line.loc[38]
net.line = net.line.drop(38)
line2 = net.line.loc[1]
net.line = net.line.drop(1)

#Add multiple lines based on formerly droped lines
pp.create_lines(net,
from_buses=[line2['from_bus'], line1['from_bus']],
to_buses=[line2['to_bus'], line1['to_bus']],
length_km=[line2['length_km'], line1['length_km']],
std_type=[line2['std_type'], line1['std_type']],
name=[line2['name'], line1['name']],
df=[line2['df'], line1['df']],
parallel=[line2['parallel'], line1['parallel']],
in_service=[line2['in_service'], line1['in_service']])

#see if powerflow runs smoothly
pp.runpp(net)

#manually check, if the lines are added correctly
check = net.line

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 75.59%. Comparing base (f1fd3d6) to head (0a69173).
Report is 2059 commits behind head on develop.

Additional details and impacted files

@@             Coverage Diff             @@
##           develop    #1403      +/-   ##
===========================================
- Coverage    75.61%   75.59%   -0.03%     
===========================================
  Files          185      185              
  Lines        21775    21772       -3     
===========================================
- Hits         16466    16458       -8     
- Misses        5309     5314       +5     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[rbolgaryn]

what is the reason for this?

[elodin]

User-friedlyness: even if the user only wants to create one line (as one is also a multiple of lines) and only passes the index as an int instead of a list, the function will work.

Also this could be a way to reduce the number of functions the user has to know by possibly replacing create_line().

What do you think about it?

[rbolgaryn]

I think we should not having this kind of logic here. If some types in pp std_type library are missing g_us_per_km, we should change it there (even if to 0).

[rbolgaryn]

The user should be responsible for providing the relevant parameters in std_type, I think

[elodin]

I agree. The user should be responsible. But the parameter should also be a natural part of the std_type, so the user knows what info needs to be provided in the first place.
Is the std_type going through a similar function as create_empty_network() during creation/loading of the net? Would make sense to add the parameter there, if it's missing.

[rbolgaryn]

Hi @elodin ,

I think it is a good idea to combine functions for 1 line and multiple lines.
I think it should happen in a different Pull Request though. I suggest to discuss this point in our meeting.

For the second point, you are right that the std_types should be adjusted accordingly, so that we don't need to have the if-clause in line 1990. I think in this PR it should be adjusted in the std_type library and then the missing parameters should raise an exception.

Roman