Okay, so yesterday I was messing around with some data from Colón, Panama. Figured I’d share how I went about it, kinda like a brain dump, ya know?
First thing I did was, I grabbed some publicly available data online. It was a CSV file with all sorts of info about businesses in Colón. I’m talking names, addresses, what they sell, all that jazz. Standard stuff, right?
Then, I fired up Python. Yeah, yeah, I know, super original. But honestly, it’s just the easiest way to wrangle data quickly. I used Pandas, naturally, to load the CSV into a DataFrame. I’m not gonna lie, there were some encoding issues at first. Had to mess around with `encoding=’latin-1’` and `encoding=’utf-8’` until it looked right. You know how it goes.
Cleaning time! The data was kinda messy. Some columns had missing values, others had weird characters. So, I went through each column, one by one, and cleaned it up. Used `fillna()` to fill in missing values with something reasonable (like “Unknown” or “N/A”). And I used `*()` to get rid of those pesky characters. It was tedious, but gotta do what you gotta do.
After cleaning, I started doing some basic analysis. What kind of businesses are most common in Colón? I used `value_counts()` on the “business type” column to find out. Turns out there are a lot of restaurants and retail stores. Not exactly shocking, but good to confirm.
Next, I wanted to visualize the data. I’m no data viz expert, but I can throw together a decent bar chart. I used Matplotlib (again, super original) to create a chart showing the distribution of business types. It wasn’t pretty, but it got the point across.
Then I thought, “Hey, let’s try to map this data!” I had the addresses, so I figured I could geocode them and plot them on a map. Geocoding is basically turning an address into latitude and longitude coordinates. I used the Geopy library for that. It took a while, because geocoding can be slow, especially with a lot of addresses. But eventually, I got the coordinates for most of the businesses.
Finally, I used Folium to create an interactive map. I plotted each business as a marker on the map. You could click on the marker to see the business name and address. It was pretty cool to see the data come to life like that.
- Here’s a quick rundown of the steps I took:
- Grabbed the data.
- Loaded it into Pandas.
- Cleaned the data (missing values, weird characters).
- Analyzed the business types.
- Created a bar chart.
- Geocoded the addresses.
- Plotted the businesses on a map using Folium.
What I Learned
Honestly, it was a pretty straightforward project. But it was a good reminder of how important data cleaning is. Garbage in, garbage out, as they say. I also learned a bit more about geocoding and how to use Folium. Definitely gonna use that library again in the future.
That’s pretty much it. Nothing groundbreaking, but hopefully someone finds this helpful. Maybe it’ll inspire you to mess around with some data from your own city or town. It’s surprisingly fun!
