[FIXED] Handling datetime object in a plot

Issue

I am trying to calculate the means per month and year and plot it in a seaborn heatmap:

           city_id    NO2     O3  season
basetime                                
2015-01-01   AQ003   7.80  63.02  winter
2015-01-02   AQ003  18.51  52.76  winter
2015-01-03   AQ003  36.52  21.13  winter
2015-01-04   AQ003  31.92  29.80  winter
2015-01-05   AQ003  11.53  56.84  winter

So this was my idea:

year = airquality_athen.index.year.rename("year")
month = airquality_athen.index.month.rename("month")

monthly_means_per_year = airquality_athen.groupby([year, month]).mean()


seven_point_five = monthly_means_per_year.pivot("month", "year", "NO2")

ax = sns.heatmap(seven_point_five)

So the problem is that i don't know how to handle datetime objects when it comes to using them in a plot. (I am getting key error: 'month') I wanted it to look like the fourth plot on this:

https://seaborn.pydata.org/generated/seaborn.heatmap.html#seaborn.heatmap

Solution

Try this:

df.index = pd.to_datetime(df.index)

df['year'] = df.index.year
df['month'] = df.index.month

monthly_means_per_year = df.groupby(['year', 'month'], as_index=False).mean()
seven_point_five = monthly_means_per_year.pivot("month", 'year', "NO2")

ax = sns.heatmap(seven_point_five)

Answered By - Rabinzel