Thread by @reuvenmlerner

Want a #Python #pandas data frame with all Apollo missions, indexed by date?

df = pd.read_html('https://t.co/1OfUmAGe6N')[2]

df['Date'] = pd.to_datetime(df['Date'].str.replace('(–.+)?,', '', regex=True))

df = df.set_index('Date')

The first line scrapes the Wikipedia page for the Apollo program, putting all HTML tables into data frames. The missions are in the third table, aka index 2.

The second line turns lines containing date ranges into single (launch) dates, also removing commas and hyphens.

That second line then takes the resulting cleaned-up date strings, and passes them to pd.to_datetime. The resulting datetime series is then assigned back to df['Date'].

Finally, we set the Date column to be df's index. And -- voila! -- now we have a time series, on which we can run time-related queries. For example, how many missions were there in every 6-month period?

df.resample('6M').count()

I'm always amazed by how easy pandas makes this.