What I want to do is put the casting in context. How old would Taylor-Johnson be when his first movie premieres and how would that compare to previous Bond actors?<\/p>\n I scraped the List of James Bond Films<\/em><\/a> Wikipedia page and dumped the data into a CSV file. Columns include movie release date, film title, and starring actor. You can download the dataset at the bottom of this post.<\/p>\n First, load the data into a pandas DataFrame.<\/p>\n We know movie release dates but we’re really<\/em> interested in the starring actor’s age on that day. We can use each actor’s birthday to create an age<\/em> column.<\/p>\n Granted, this isn’t a perfect way to calculate a time delta in units of years. Dividing days by 365.25 won’t account for leap year exceptions on the century, or exceptions to the exceptions on millennia, or leap seconds, or relativistic time travel effects. But fortunately there won’t be enough pixels in our image for that to be a issue. And, of course, movies are filmed some time in advance of release, so this value won’t reflect their exact on-screen age. But if movies can bend the truth then so can we. At this point the DataFrame’s I’m using a custom Matplotlib style that’s a forked version of the built-in ggplot<\/em> style, which in turn is a play on R’s ggplot2<\/a>. You can download it at the bottom of this post.<\/p>\n The general idea is to plot a line showing Bond’s age over time, including discontinuities when he’s re-cast, and overlay markers on the line to represent each movie in the series. Since we want the scatter plot to be layered above the line, we pass a layer argument to For the scatter plot, we’ll loop through each actor and use their name to create a DataFrame view<\/em>, i.e. a subset of rows. For example, Pierce Brosnan starred in four Bond films, so At this point we’ve plotted the data and it’s a matter of final adjustments.<\/p>\n A handy way to create a series of x-axis datetime values is to use I leave a little extra room above the top y-tick for a legend.<\/p>\n In defining a legend, Finally, set a title and save the figure. I bump ![\"\"](\"https:\/\/wollen.org\/blog\/wp-content\/uploads\/2024\/03\/ATJ_bullet-train2.gif\")
<\/a>
\n1. Prepare the data.<\/h4>\n
import pandas as pd\r\nimport matplotlib.pyplot as plt\r\nfrom matplotlib.dates import DateFormatter\r\n\r\ndf = pd.read_csv(\"bond_data.csv\", parse_dates=[\"date\"])<\/pre>\n
apply<\/code> a function along the horizontal axis.<\/p>\nbirthday_dict = {\"Sean Connery\": \"August 25, 1930\",\r\n \"George Lazenby\": \"September 5, 1939\",\r\n \"Roger Moore\": \"October 14, 1927\",\r\n \"Timothy Dalton\": \"March 21, 1946\",\r\n \"Pierce Brosnan\": \"May 16, 1953\",\r\n \"Daniel Craig\": \"March 2, 1968\"}\r\n\r\ndf.loc[:, 'age_at_release'] = df.apply(lambda row: (row['date'] - pd.Timestamp(birthday_dict[row['actor']])).days \/ 365.25, axis=1)<\/pre>\nhead()<\/code> looks like this:<\/p>\n date title actor age_at_release\r\n0 1962-10-05 Dr. No Sean Connery 32.112252\r\n1 1963-10-10 From Russia with Love Sean Connery 33.125257\r\n2 1964-09-17 Goldfinger Sean Connery 34.064339\r\n3 1965-12-09 Thunderball Sean Connery 35.290897\r\n4 1967-06-12 You Only Live Twice Sean Connery 36.796715<\/pre>\n
\n2. Plot the data.<\/h4>\n
zorder<\/code>.<\/p>\nplt.style.use(\"wollen_ggplot.mplstyle\")\r\nfig, ax = plt.subplots()\r\n\r\nax.plot(df['date'], df['age_at_release'], color=\"black\", linewidth=0.6, zorder=1)<\/pre>\n
df2<\/code> will have four rows and ax.scatter()<\/code> will add four markers to the figure. Set zorder<\/code> to 2 so markers are drawn on the topmost layer.<\/p>\nfor name in df['actor'].unique():\r\n df2 = df.copy()\r\n df2 = df2[df2['actor'] == name]\r\n ax.scatter(df2['date'], df2['age_at_release'], label=name, zorder=2)<\/pre>\n
pd.date_range<\/code>. The “5YS” argument passed to freq<\/code> requests a value every five years located at Y<\/strong>ear S<\/strong>tart (January 1st).<\/p>\nx_ticks = pd.date_range(pd.Timestamp(\"Jan 1 1960\"), pd.Timestamp(\"Jan 1 2025\"), freq=\"5YS\")\r\nax.set_xticks(x_ticks)\r\nax.xaxis.set_major_formatter(DateFormatter(\"%Y\"))\r\nax.set_xlim(pd.Timestamp(\"July 1 1958\"), pd.Timestamp(\"July 1 2026\"))\r\n\r\ny_ticks = range(0, 65, 5)\r\nax.set_yticks(y_ticks)\r\nax.set_ylim(0, 64)\r\nax.set_ylabel(\"Age at Release (Years)\")<\/pre>\n
ncol<\/code> refers to number of columns. I want the legend to be a single row at the top of the axis.<\/p>\ndpi<\/code> slightly to improve readability.<\/p>\nax.legend(loc=\"upper center\", ncol=6)\r\n\r\nax.set_title(\"James Bond Actors | Age\")\r\n\r\nplt.savefig(\"bond_actor_age.png\", dpi=120)<\/pre>\n
\n3. The output.<\/h4>\n
![\"\"](\"https:\/\/wollen.org\/blog\/wp-content\/uploads\/2024\/03\/bond_actor_age.png\")
<\/a><\/p>\n