Going for Gold: Visualizing Olympic Medal Counts with SAS Viya
As a (very) amateur skier, I spent a lot of time last month watching the winter Olympics in Milan and seeing athletes do flips and twists that I had never even realized were possible. Now that the Olympics are all wrapped up, it’s my time to shine instead – but instead of quad axels and gold medals, I’ll use tables and graphs to break down medal distribution across countries. In this article, I’ll walk through an entire data science pipeline using SAS Viya, from accessing web data using Python, to creating maps with PROC SGMAP.
Scraping HTML using Python
The first thing I need to do is find a data source. I’ll use the 2026 Winter Olympics medal table Wikipedia page, which looks like this:
I need to scrape this HTML table and convert it to a SAS data set. Luckily, I can use the read_html() function from the Pandas library in Python to do this in one simple step, since SAS supports Python programming directly from SAS Studio or Visual Studio Code.
You can find all of the code used in the post here.
...
import requests
import pandas as pd
import saspy
sas=saspy.SASsession()
url = "https://en.wikipedia.org/wiki/2026_Winter_Olympics_medal_table"
# Add a browser-like user agent to avoid 403
headers = {
"User-Agent": (
"Mozilla/5.0 (Windows NT 10.0; Win64; x64) "
"AppleWebKit/537.36 (KHTML, like Gecko) "
"Chrome/119.0.0.0 Safari/537.36"
)
}
# Fetch page with headers
response = requests.get(url, headers=headers)
response.raise_for_status()
# Pandas reads the HTML from the response content
tables = pd.read_html(response.text)
# The medal table is the first table
df = tables[2]
print(df)
# Send to SASUSER library as SASUSER.MEDALS
sas.df2sd(df, table='medals', libref='sasuser')
print("SASUSER.MEDALS created successfully.")
...
Note that if you are following along, the read_html() function requires the lxml library to process HTML, which you may need to have installed in your environment.
This scrapes the HTML table and creates a new table called MEDALS in the SASUSER library:
Cleaning the Data
However, the new table from Wikipedia isn’t quite ready to use yet. There are a few rows that will cause problems when I try to visualize them, such as a Totals row and the Individual Neutral Athletes row. There are also some quirks: Italy is written as Italy*. Finally, in order to create a map, I need to assign a standard ID key to join it with my existing map data. I’ll use the ISO two-character country code for this.
data work.medal_clean;
set sasuser.medals;
if index(upcase(NOC), 'TOTALS') then delete;
if NOC =: 'Individual Neutral' then delete;
NOC = compress(NOC, '*');
select (strip(NOC));
when ('Norway') ISO2='NO';
when ('United States') ISO2='US';
…
otherwise ISO2='';
end;
run;
*Sort the data by number of gold medals;
proc sort data=work.medal_clean out=medal_clean;
by descending gold;
run;Visualizing Medal Counts
Now that the data is cleaned and ready to go, the last step is to visualize it. I’ll create a map using PROC SGMAP and a bar chart using PROC SGPLOT. SAS comes with a number of map data sets already available in the MAPSGFK library, so creating a choropleth map is as simple as identifying which one you need. In this case, I want to create a world map, so I’ll use the MAPSGFK.WORLD table.
title "Total Medals: Milan Cortina 2026";
proc sgmap mapdata=mapsgfk.world
maprespdata=work.medal_clean;
choromap total / mapid=id id=iso2 colormodel=(white grey);
run;
title;
proc sgplot data=work.medal_clean(obs=10);
hbar noc / response=gold
datalabel
fillattrs=(color=gold);
yaxis label="Nation" discreteorder=data;
xaxis label="Gold Medals";
title "Top 10 Nations - Gold Medals, 2026 Winter Olympics";
run;
The result is a map:
And a bar chart:
Conclusion
It turns out that scraping medal tables is much easier than actually winning one—who knew? Luckily, SAS is far more forgiving than a panel of Olympic judges, so even a few missteps along the way won’t cost us any points.
By combining Python for data collection and SAS procedures for cleaning and visualization, we’re able to move from raw HTML to polished insights in just a few steps, all on the SAS Viya platform. This same workflow applies to countless real‑world scenarios where data lives online and needs to be transformed quickly into something meaningful. And while the athletes are training for 2030, we can keep exploring new ways to turn raw data into gold‑medal‑worthy analytics using SAS.
