Need to reload input file when updating/plotting in a loop?

[paulf81]

Bug description
Had a code set up to change FLORIS settings and plot in a loop and discovered it only worked if I reopened the input file (or if don't plot). Is this expected operation

To Reproduce
Made this code example, it doesn't work for me unless I:
1). Uncomment the code which reopens the model from input yaml
2) Dont plot

import matplotlib.pyplot as plt
import numpy as np

from floris.tools import FlorisInterface
from floris.tools.visualization import visualize_cut_plane


# Inputs
X = [0, 500]
Y = [0, 0]
yaw_angles = np.zeros((1, 1, 2)) # 1 wd/ 1ws/ N*N turbines
fi_gch = FlorisInterface("inputs/gch.yaml")


for wd_idx, wind_direction in enumerate([270.,  280.]):

    # Code works if this lined is uncommented
    # fi_gch = FlorisInterface("inputs/gch.yaml") 


    fi_gch.reinitialize(  layout=( X, Y ), wind_speeds=[8.0], wind_directions=[wind_direction], turbulence_intensity=0.06 )
    fi_gch.calculate_wake(yaw_angles=yaw_angles)
    turbine_powers_gch = fi_gch.get_turbine_powers()
    turbine_powers_gch = np.sort(turbine_powers_gch.flatten())[::-1]/1000.

    # Code also works if comment out this block
    fig, ax = plt.subplots()
    horizontal_plane_gch = fi_gch.get_hor_plane(x_resolution=100, y_resolution=100)
    visualize_cut_plane(horizontal_plane_gch, ax=ax, title="Wind Direction = %.1f" % wind_direction)


plt.show()

Expected behavior
Was expecting the code to work without the call back to the yaml, is something in plotting "sticking"?

Screenshots, if applicable
If the code works plots:

Floris Version
V3

[bayc]

Yea, I noticed this too. My thought was that something was sticking from the grid used for the flow field visualizations, but haven't been able to dig into it yet.

[rafmudaf]

@paulf81 can you confirm that this is fixed by #327?

[paulf81]

Confirmed! I had to update the function call from get_ to calculate_ but then I modified the test also a little bit to confirm that after going back and forth a few times I produce the same answer, and have the same total number of grid points. I think all confirmed via:

import matplotlib.pyplot as plt
import numpy as np

from floris.tools import FlorisInterface
from floris.tools.visualization import visualize_cut_plane


# Inputs
X = [0, 500]
Y = [0, 0]
yaw_angles = np.zeros((1, 1, 2)) # 1 wd/ 1ws/ N*N turbines
fi_gch = FlorisInterface("inputs/gch.yaml")
fi_gch.reinitialize(  layout=( X, Y ))
print('starting shape:', fi_gch.floris.grid.x.shape)
for wd_idx, wind_direction in enumerate([270.,  270., 280., 270.]):

    # Code works if this lined is uncommented
    # fi_gch = FlorisInterface("inputs/gch.yaml") 


    fi_gch.reinitialize(  layout=( X, Y ), wind_speeds=[8.0], wind_directions=[wind_direction], turbulence_intensity=0.06 )
    fi_gch.calculate_wake(yaw_angles=yaw_angles)
    turbine_powers_gch = fi_gch.get_turbine_powers()
    turbine_powers_gch = np.sort(turbine_powers_gch.flatten())[::-1]/1000.

    # Code also works if comment out this block
    fig, ax = plt.subplots()
    horizontal_plane_gch = fi_gch.calculate_horizontal_plane(x_resolution=100, y_resolution=100)
    visualize_cut_plane(horizontal_plane_gch, ax=ax, title="Wind Direction = %.1f" % wind_direction)

    print(turbine_powers_gch, fi_gch.floris.grid.x.shape)


plt.show()

yielding:

starting shape: (1, 1, 2, 3, 3)
[1691.32648289  361.23204874] (1, 1, 2, 3, 3)
[1691.32648289  361.23204874] (1, 1, 2, 3, 3)
[1691.32648289 1290.71622033] (1, 1, 2, 3, 3)
[1691.32648289  361.23204874] (1, 1, 2, 3, 3)

and working plots, all good!

[rafmudaf]

Currently, the FlorisInterface.calculate_XX_plane() routines reinitialize Floris with a single atmospheric condition (required for plotting). While we reset the solver settings, we do not reset the state of Floris prior to making the change for plotting.