[Python] Get aggregates from different Dataframe to current Dataframe with conditions

I have a harvest dataframe and a weather dataframe. I want to get the number of days above a temp threshold for the previous x months before harvest for all blocks. Note the harvest dataframe includes multiple years and the id is not 1-1 between frames, ie 2 blocks in harvest df can share an ID that correspond to a location in the weather frame.

My current (working) code is below, but it is VERY slow, on the order of minutes. I want to speed it up but unclear how.

def days_above_thresh(x, weather_df):
return weather_df.loc[
(weather_df["id"]==x.id) & \
(weather_df["day"]>=x['harvest_date']-DateOffset(months=2)) & \
(weather_df["day"]<=x['harvest_date']) & \
(weather_df["temperature_max"]>30),
"temperature_max"].count()

harvest_df["days_above_30"] = harvest_df.apply(days_above_thresh , args=(weather_df,), axis=1)


The dataframes would look something like this -

weather_df
id day temperature_max
1 2020-01-01 30
1 2020-01-02 32
1 2020-01-03 28
1 2020-01-04 25
.
.
.
2 2020-01-01 10
2 2020-01-02 15
2 2020-01-03 17
2 2020-01-04 12
.
.
.

harvest_df
id farm_id harvest_date
1 87 2020-01-02
1 86 2020-01-03
2 13 2020-01-30

Continue reading...