57 — Group: data by category

Counting (guides 36 and 54) answers how many; grouping answers how many per category: it splits one list of dicts into lists, one per value of a key.

#Steps

1. Create data.json, a list of dicts — each has a name, category, price:

   [
     {"name": "Mina", "category": "fruit", "price": 3},
     {"name": "Jun", "category": "veggie", "price": 2},
     {"name": "Sora", "category": "fruit", "price": 5},
     {"name": "Tom", "category": "veggie", "price": 1},
     {"name": "Ari", "category": "fruit", "price": 4}
   ]
   

   copyrun it →

2. Load it with json_load from guide 14. The manual group uses the if word in tally idea from guide 36 — a new category starts an empty list:

   use file latest
   items = json_load("data.json")
   groups = {}
   for item in items:
       cat = item["category"]
       if cat not in groups:
           groups[cat] = []
       groups[cat].append(item)
   

   copyrun it →

3. setdefault does the "new key gets an empty list" step in one call, then append always works — the shorter spelling of the same loop:

   quick = {}
   for item in items:
       quick.setdefault(item["category"], []).append(item)
   

   copyrun it →

4. The full program groups, reports counts and totals, and proves setdefault agrees. Save it as group.nme:

   # group.nme — group a list of dicts by category, then report counts.
   # Run: nme r group
   
   use file latest
   items = json_load("data.json")
   show "loaded " + str(len(items)) + " items"
   
   groups = {}
   for item in items:
       cat = item["category"]
       if cat not in groups:
           groups[cat] = []
       groups[cat].append(item)
   
   show "counts per category:"
   for cat in groups:
       show cat + ": " + str(len(groups[cat])) + " items"
   
   show "total price per category:"
   for cat in groups:
       total = 0
       for item in groups[cat]:
           total = total + item["price"]
       show f"{cat}: ${total}"
   
   quick = {}
   for item in items:
       quick.setdefault(item["category"], []).append(item)
   
   show "setdefault agrees:"
   for cat in quick:
       show f"{cat}: {len(quick[cat])} items"
   

   copyrun it →

   The outer loop walks categories, the inner loop walks each category's items.

5. Run it with data.json in the folder:

   nme r group
   

   copy

   loaded 5 items
   counts per category:
   fruit: 3 items
   veggie: 2 items
   total price per category:
   fruit: $12
   veggie: $3
   setdefault agrees:
   fruit: 3 items
   veggie: 2 items
   

   copy

   Each category is a list, so the report can count it, name it, or add prices.

#Try it yourself

Group by price instead of category — every price is its own group, showing which items cost the same.

#What you learned

• Grouping splits one list of dicts into a dict of lists by a key's value.
• if cat not in groups starts a new list; groups[cat].append(item) fills it.
• groups.setdefault(cat, []).append(item) is the same in one call.
• The outer loop walks categories; the inner loop walks each category's items.