78 — Data: analyzing a month of temperatures

Averages are a start, not the whole story. This guide analyzes a real-ish dataset: 30 daily temperatures. The numbers describe it, the histogram draws its shape, and a report file keeps the findings — the same three steps a data analysis project does at any size.

#Steps

1. Create temps.json with 30 daily high temperatures (Celsius):

   [22, 24, 21, 25, 27, 26, 23, 20, 19, 24,
    26, 28, 29, 30, 27, 25, 24, 23, 22, 26,
    28, 31, 32, 29, 26, 24, 23, 21, 20, 25]
   

   copyrun it →

2. Load it with json_load (guide 14) and compute the statistics from guide 54:

   use file latest
   
   from statistics import mean, median
   
   temps = json_load("temps.json")
   average = round(mean(temps), 1)
   middle = median(temps)
   hottest = max(temps)
   coolest = min(temps)
   

   copyrun it →

   round(mean(temps), 1) keeps one decimal, so the report does not print 25.333333333333332.

3. A histogram shows the shape of the data: count how many days fall into each range, then draw one bar per range with string multiplication (guide 71):

   ranges = ["under 20", "20-24", "25-29", "30+"]
   counts = [0, 0, 0, 0]
   for t in temps:
       if t < 20:
           counts[0] = counts[0] + 1
       elif t < 25:
           counts[1] = counts[1] + 1
       elif t < 30:
           counts[2] = counts[2] + 1
       else:
           counts[3] = counts[3] + 1
   
   for i in range(4):
       show f"{ranges[i]:8s} {'#' * counts[i]}"
   

   copyrun it →

   The elif chain places each day in exactly one range, so the four counts always add up to 30 — a check worth making.

4. The full program also writes the findings into report.txt. Save analysis.nme:

   # analysis.nme — analyze a month of temperatures and write a report.
   # Run: nme r analysis
   # The file temps.json must exist in the same folder.
   
   use file latest
   
   from statistics import mean, median
   
   temps = json_load("temps.json")
   average = round(mean(temps), 1)
   middle = median(temps)
   hottest = max(temps)
   coolest = min(temps)
   
   lines = []
   lines.append(f"temperature report ({len(temps)} days)")
   lines.append(f"  average: {average}")
   lines.append(f"  median:  {middle}")
   lines.append(f"  hottest: {hottest}")
   lines.append(f"  coolest: {coolest}")
   
   ranges = ["under 20", "20-24", "25-29", "30+"]
   counts = [0, 0, 0, 0]
   for t in temps:
       if t < 20:
           counts[0] = counts[0] + 1
       elif t < 25:
           counts[1] = counts[1] + 1
       elif t < 30:
           counts[2] = counts[2] + 1
       else:
           counts[3] = counts[3] + 1
   
   lines.append("distribution:")
   for i in range(4):
       lines.append(f"  {ranges[i]:8s} {'#' * counts[i]}")
   
   report = "\n".join(lines)
   show report
   file_write("report.txt", report)
   show "saved report.txt"
   

   copyrun it →

   Every finding is appended to lines first, then printed and saved with the same text — the report cannot drift between screen and file.

5. Run it:

   nme r analysis
   

   copy

   temperature report (30 days)
     average: 25
     median:  25.0
     hottest: 32
     coolest: 19
   distribution:
     under 20 #
     20-24    #############
     25-29    ##############
     30+      ##
   

   copy

   The numbers and the histogram agree: most days sit between 20 and 29, with one cool day and two hot ones.

#Try it yourself

Add a second file with last year's temperatures and compare the averages, or count "hot days" (>= 30) and put that number into the report too. The ranges list is data — try ["under 18", "18-27", "28+"] and watch the elif chain still place every day.

#What you learned

• One elif chain places every value into exactly one range.
• A histogram is just counts per range drawn with '#' * count.
• round(value, 1) keeps report numbers short.
• Building the report as a lines list keeps screen and file identical.
• Statistics plus a histogram together describe a dataset better than either alone.