What's Wrong With My Vdims In My HoloViews Python Code?

display(df_top5_frac.head()) The code below produces an error. %opts Overlay [width=800 height=600 legend_position='top_right'] Curve hv.Curve((df_top5_frac['Blocked Driveway'])

Solution 1:

Cause of your error:
The vdim should be the name of the column you want to have on the y-axis, but the column name 'Fraction' doesn't exist, so you get the error.

Here's a possible solution:
When you set hour as the index, you could specify: kdim='hour' and vdim='blocked_driveway', but in this case you don't really need them and can leave them out:

# import libraries
import numpy as np
import pandas as pd
import holoviews as hv
hv.extension('bokeh')

# create sample data
data = {'hour': ['00', '01', '02'],
        'blocked_driveway': np.random.uniform(size=3),
        'illegal_parking': np.random.uniform(size=3),
        'street_condition': np.random.uniform(size=3),}

# create dataframe and set hour as index
df = pd.DataFrame(data).set_index('hour')

# create curves: 
# in this case the index is automatically taken as kdim
# and the series variable, e.g. blocked_driveway is taken as vdim
plot1 = hv.Curve(df['blocked_driveway'], label='blocked_driveway')
plot2 = hv.Curve(df['illegal_parking'], label='illegal_parking')
plot3 = hv.Curve(df['street_condition'], label='street_condition')

# put plots together
(plot1 * plot2 * plot3).opts(legend_position='top', width=600, height=400)

Alternative and shorter solution:
In this case however I would use library hvplot which is built on top of holoviews.
It has even easier syntax and you need a lot less code to get the plot you want:

import hvplot.pandas

# you don't have to set hour as index this time, but you could if you'd want to.
df.hvplot.line(
    x='hour', 
    y=['blocked_driveway', 
       'illegal_parking',
       'street_condition'],
)

Resulting plot: