Unknown df in get_farm_AEP_wind_rose_class

[JarneVB]

I tried calculating the farm AEP with a wind rose by using get_farm_AEP_wind_rose_class, but I get the following error message:

'DataFrame' object has no attribute 'df'

It is about these lines:

   wind_speeds_wind_rose = wind_rose.df.ws.unique()
   wind_directions_wind_rose = wind_rose.df.wd.unique()
   self.reinitialize(wind_speeds=wind_speeds_wind_rose, wind_directions=wind_directions_wind_rose)
   # Build the frequency matrix from wind rose
   freq = wind_rose.df.set_index(['wd','ws']).unstack().values`

It looks like df is defined nowhere. How can I fix this?

[paulf81]

hi @JarneVB, I would like to help. Just to check something quickly, do you get this error if you run the example script: 08_calc_aep_from_rose_use_class.py ?

[JarneVB]

Thank you for your fast response! The example script works fine, it's just my own script. Might it have something to do with the wind rose argument that I use as input? I used a wind rose generated by make_wind_rose_from_weibull().

[paulf81]

Would it be possible for you to post a minimal version of the script that produces the error? I'd be happy to debug

[JarneVB]

from floris.floris.tools.floris_interface import FlorisInterface 
from floris.floris.tools import FlorisInterface, wind_rose, WindRose
D = 112
X_farm = [1*D, 10*D, 20*D]
Y_farm = [1*D, 10*D, 20*D]

fi_weibull_jensen = FlorisInterface("floris/examples/inputs/jensen.yaml")
fi_weibull_jensen.reinitialize(layout_x=X_farm, layout_y=Y_farm)
wind_data_weibull =  make_wind_rose_from_weibull()
jensen_aep = fi_weibull_jensen.get_farm_AEP_wind_rose_class(wind_rose=wind_data_weibull)
print("Farm AEP (default options): {:.3f} GWh".format(jensen_aep / 1.0e9))

I reinitialize for a farm of 50 turbines, but it also gives the same problem for less turbines, so this is a bit shorter.

[paulf81]

Hi! Ok, I think there is an issue in our code, the function make_wind_rose_from_weibull() both returns a dataframe, and updates the member dataframe of the WindRose class. We have a TODO item to remove the redundant return which I'll convert into an issue now. In the meanwhile, this code works for me (ignoring the returned df and using the updated wind_rose object)

from floris.tools.floris_interface import FlorisInterface 
from floris.tools import FlorisInterface, WindRose

D = 112
X_farm = [1*D, 10*D, 20*D]
Y_farm = [1*D, 10*D, 20*D]

fi_weibull_jensen = FlorisInterface("inputs/jensen.yaml")
fi_weibull_jensen.reinitialize(layout_x=X_farm, layout_y=Y_farm)

wind_rose = WindRose()
_ =  wind_rose.make_wind_rose_from_weibull()

jensen_aep = fi_weibull_jensen.get_farm_AEP_wind_rose_class(wind_rose=wind_rose)
print("Farm AEP (default options): {:.3f} GWh".format(jensen_aep / 1.0e9))

[JarneVB]

This works!! Thank you for the help!

[JarneVB]

One more question: get_farm_AEP_wind_rose_class doesn't have an argument 'turbine_weights'. Is there another option to use the wake of neighboring wind farms, but excluding them from the AEP calculation?

[paulf81]

Yes, I think this just needs to turbine_weights added to the function get_farm_AEP_wind_rose_class, I'll open an issue

[paulf81]

Ok I've opened issue #540