Plotly: Pass array to 'fillcolor' argument - colors

Question

I am trying to pass a array of colors to the 'fillcolor' argument to conditionally color the area below a graph.
First, passing a string with the color or rgba code to the 'fillcolor' argument works perfectly fine:
fig.add_trace(
go.Scatter(x=x,
y=y,
mode='lines',
fill='tozeroy',
fillcolor='red',
line=dict(color='rgba(255, 0, 0, 0.9)')))
When passing an array to the 'fillcolor' argument, such as
fig.add_trace(
go.Scatter(x=x,
y=y,
name='price',
mode='lines',
fill='tozeroy',
fillcolor=np.where(y > -5, 'green', 'red'),
line=dict(color='rgba(255, 0, 0, 0.9)')))
leads to this error:
ValueError:
Invalid value of type 'numpy.ndarray' received for the 'fillcolor' property of scatter
Also, passing an array directly or passing 'fillcolor=dict(color=array)', does not work.
Is there a chance to pass an array of colors and hence fill the area to the axis conditionally?
Thanks in advance!

Answer 1

as per comments fillcolor is not an array - it is an attribute of a trace
to achieve a multi-color figure you require a trace for each color line (filled line) required in figure
have uses pandas and Plotly Express in this solution, using your original code to set it up
this does give a marginally different shape plot due to way tozeroy works
import numpy as np
import plotly.graph_objects as go
x = np.linspace(0, 20, 100)
y = np.random.randint(-7, 3, 100)
fig = go.Figure()
fig.add_trace(
go.Scatter(
x=x,
y=y,
mode="lines",
fill="tozeroy",
fillcolor="red",
line=dict(color="rgba(255, 0, 0, 0.9)"),
)
)
fig.show()
# create a dataframe to simplify solution. green and red lines logic as additional columns
df = pd.DataFrame({"x": x, "y": y})
df = df.assign(
y_green=np.where(df["y"] > -5, df["y"], 0), y_red=np.where(df["y"] > -5, 0, df["y"])
)
# create the plot
fig = px.line(
df, x="x", y=["y_green", "y_red"], color_discrete_sequence=["green", "red"]
).for_each_trace(lambda t: t.update(fillcolor=t.line.color, fill="tozeroy"))
fig