import matplotlib.pyplot as plt

# Data
labels = ["bg","cs","fi","id","it","nl","nb","pl","pt","th","tr","zh","ja","ru","de","sh","sr",
          "ab","ace","af","gsw","arz","ast","as","azb","az","bar","sgs","ba","be","be-tarask",
          "bho","bjn","bn","bo","bs","bug","cdo","ce","ckb","da","diq","dty","el","et","eu",
          "fo","frr","gan","gl","gu","hak","hif","hi","hr","hy","ilo","is","jv","kk","km",
          "kn","ksh","ku","lad","la","lb","lez","lmo","lt","lv","mai","map-bms","mhr","min",
          "mk","ml","mrj","mr","ms","myv","my","mzn","nah","nap","nds","nds-nl","ne","nn","nv",
          "os","pa","pnb","ps","rue","sa","sco","se","si","sk","sl","so","sq","su","szl","ta",
          "te","tg","tk","udm","ur","uz","vec","wa","wuu","yi","yo","lzh","nan","yue","en","ar",
          "ca","es","fa","he","ko","ro","uk","vi"]
sizes = [415,1103,635,3563,1425,1244,395,3258,3587,584,1018,6286,1581,7504,9985,128,1471,
         2,19,125,50,70,40,91,28,496,9,21,70,175,87,34,77,894,22,202,15,6,31,89,168,48,15,
         9,226,289,14,20,24,210,75,22,31,518,434,914,13,70,48,152,48,430,18,40,11,69,41,23,
         9,267,69,49,8,5,293,87,422,14,342,234,8,70,23,17,5,50,6,150,147,19,12,98,3,8,21,143,
         45,12,84,372,267,2,92,317,40,919,290,46,33,9,32,61,11,6,77,22,8,66,39,186,44205,5973,
         732,3692,2495,740,616,372,1254,1161]

# Explode largest slice (English) slightly
explode = [0.02 if lbl == "en" else 0 for lbl in labels]

# Figure
fig, ax = plt.subplots(figsize=(12,12))
# For readability, show only top N slices with labels; group the rest as "Other"
N = 15
# get indices of top N by size
top_idx = sorted(range(len(sizes)), key=lambda i: sizes[i], reverse=True)[:N]
top_labels = [labels[i] for i in top_idx]
top_sizes = [sizes[i] for i in top_idx]
other_size = sum(sizes) - sum(top_sizes)
if other_size > 0:
    top_labels.append("Other")
    top_sizes.append(other_size)
    top_explode = [0.1 if lab=="en" else 0 for lab in top_labels]
else:
    top_explode = [0.1 if lab=="en" else 0 for lab in top_labels]

# Pie chart
wedges, texts, autotexts = ax.pie(
    top_sizes,
    labels=top_labels,
    autopct=lambda pct: f"{pct:.1f}%\n({int(pct/100*sum(top_sizes)):,})",
    startangle=140,
    textprops={'fontsize':10}
)

# Styling
ax.axis('equal')
plt.suptitle("HotCat users by Wikipedia language version, May 2021", fontsize=14)
plt.title("(top 15 + Other) — total: 117,670", fontsize=11)
plt.tight_layout()
plt.show()