Exercise

# Finding sources of weather delays

The final step in this case study is to use the `persisted_weather_delays`

Dask DataFrame to determine the percentage of
delayed flights per weather event.

Your job is to compute the number of delayed flights by weather events and divide by the total number of flights. You'll
then use `.nlargest(5)`

to retrieve the five highest contributions to the number of delayed flights. Finally, you'll compute
the average length of the delay for the 5 leading contributions.

Instructions

**100 XP**

- Count the
`by_event['WEATHER_DELAY']`

column and divide by the total number of delayed flights from`persisted_weather_delays`

, multiply by 100 and assign the result to`pct_delayed`

. - Compute & print the five largest values of
`pct_delayed`

with`.nlargest(5)`

. - Calculate the mean of the
`by_event['WEATHER_DELAY']`

column and return the 5 largest entries with`.nlargest(5)`

. - Compute & print
`avg_delay_time`

.