Plotly Choropleth

Question

I am working with a Kaggle dataset "US Accidents" (which can be downloaded here ) that has 3 million records on traffic accident data. A quick exploration shows that California contains the most accidents. I thought a choropleth viz would be cool to implement however, the data on my Choropleth is inaccurate and was wondering where I am going wrong/how to fix it.

Here is my code...

states_by_accident = df.State.value_counts()

import plotly.graph_objects as go
fig = go.Figure(data = go.Choropleth(
    locations = df.State.unique(),
    z = states_by_accident,
    locationmode = 'USA-states',
    colorscale = 'Blues'
))


fig.update_layout(
       geo_scope = 'usa'
   )
fig.show()

I have tried converting the colors to a log scale which helped spread out the coloring but it still displayed Ohio as having the most accidents which is inaccurate.

Answer 1

This is happening because df.State.unique() doesn't have the states in the same order as the values in states_by_accident .

You can fix this by instead passing the argument locations = states_by_accident.index to go.Chloropleth so that the locations and values are consistent:

fig = go.Figure(data = go.Choropleth(
    locations = states_by_accident.index,
    z = states_by_accident,
    locationmode = 'USA-states',
    colorscale = 'Blues'
))