TweetStats Alternative: How to Graph Your Tweet Stats Now That TweetStats Is Gone

# TweetStats-style graphs for any public X account: tweet timeline, tweets per
# weekday and hour, density heatmap, top @replies, posting clients, reposted
# accounts, TweetCloud and HashCloud, saved as tweetstats_<handle>.png + .csv
import os, re, csv, requests
from collections import Counter
from datetime import datetime, timezone
import matplotlib
matplotlib.use("Agg")
import matplotlib.pyplot as plt

HEADERS = {"X-API-Key": os.environ["TWITTERAPI_IO_KEY"]}
HANDLE = "nasa"
START, END = (2025, 10), (2026, 9)  # inclusive (year, month)
STOP = set("the a an and or to of in on for is it at be this that with i you my me we our are was rt https co amp".split())

def months(start, end):
    y, m = start
    while (y, m) <= end:
        yield y, m
        y, m = (y + 1, 1) if m == 12 else (y, m + 1)

def month_bounds(y, m):
    lo = datetime(y, m, 1, tzinfo=timezone.utc)
    hi = datetime(y + (m == 12), m % 12 + 1, 1, tzinfo=timezone.utc)
    return int(lo.timestamp()), int(hi.timestamp())

def fetch_month(y, m):
    lo, hi = month_bounds(y, m)
    query = f"from:{HANDLE} since_time:{lo} until_time:{hi}"
    cursor, out = "", []
    while True:
        r = requests.get(
            "https://api.twitterapi.io/twitter/tweet/advanced_search",
            headers=HEADERS,
            params={"query": query, "queryType": "Latest", "cursor": cursor},
            timeout=20,
        )
        r.raise_for_status()
        body = r.json()
        out += body.get("tweets", [])
        if not body.get("has_next_page"):
            return out
        cursor = body["next_cursor"]

tweets = [t for y, m in months(START, END) for t in fetch_month(y, m)]
for t in tweets:
    t["_dt"] = datetime.strptime(t["createdAt"], "%a %b %d %H:%M:%S %z %Y")

per_month = Counter(t["_dt"].strftime("%Y-%m") for t in tweets)
per_weekday = Counter(t["_dt"].weekday() for t in tweets)
per_hour = Counter(t["_dt"].hour for t in tweets)
density = [[0] * 24 for _ in range(7)]
for t in tweets:
    density[t["_dt"].weekday()][t["_dt"].hour] += 1
replied_to = Counter(t.get("inReplyToUsername") for t in tweets if t.get("isReply") and t.get("inReplyToUsername"))
clients = Counter(t.get("source") or "unknown" for t in tweets)
reposted = Counter(((t.get("retweeted_tweet") or {}).get("author") or {}).get("userName")
                   for t in tweets if t.get("retweeted_tweet"))
reposted.pop(None, None)
hashtags = Counter(h["text"].lower() for t in tweets for h in (t.get("entities") or {}).get("hashtags") or [])
words = Counter(w for t in tweets if not t.get("retweeted_tweet")
                for w in re.findall(r"[a-z']{3,}", re.sub(r"https?://\S+|@\w+|#\w+", "", t.get("text", "").lower()))
                if w not in STOP)

fig, ax = plt.subplots(2, 2, figsize=(14, 9))
labels = [f"{y}-{m:02d}" for y, m in months(START, END)]
ax[0][0].bar(labels, [per_month[k] for k in labels], color="#1d9bf0")
ax[0][0].set_title("Tweet timeline (per month)")
ax[0][0].tick_params(axis="x", rotation=60)
ax[0][1].bar(["Mon", "Tue", "Wed", "Thu", "Fri", "Sat", "Sun"], [per_weekday[d] for d in range(7)], color="#1d9bf0")
ax[0][1].set_title("Tweets per weekday (UTC)")
ax[1][0].bar(range(24), [per_hour[h] for h in range(24)], color="#1d9bf0")
ax[1][0].set_title("Tweets per hour (UTC)")
im = ax[1][1].imshow(density, aspect="auto", cmap="Blues")
ax[1][1].set_yticks(range(7), ["Mon", "Tue", "Wed", "Thu", "Fri", "Sat", "Sun"])
ax[1][1].set_title("Tweet density (weekday x hour, UTC)")
fig.colorbar(im, ax=ax[1][1])
fig.suptitle(f"TweetStats-style graphs for @{HANDLE}: {len(tweets)} tweets")
fig.tight_layout()
fig.savefig(f"tweetstats_{HANDLE}.png", dpi=120)

with open(f"tweetstats_{HANDLE}.csv", "w", newline="") as f:
    w = csv.writer(f)
    w.writerow(["id", "createdAt", "isReply", "inReplyToUsername", "source", "likeCount", "retweetCount", "viewCount", "text"])
    for t in tweets:
        w.writerow([t["id"], t["createdAt"], t.get("isReply"), t.get("inReplyToUsername"), t.get("source"),
                    t.get("likeCount"), t.get("retweetCount"), t.get("viewCount"), t.get("text")])

for title, c in [("Top @replies", replied_to), ("Interface used", clients), ("Reposts of", reposted),
                 ("TweetCloud", words), ("HashCloud", hashtags)]:
    print(f"{title}: " + ", ".join(f"{k} ({v})" for k, v in c.most_common(10)))
print(f"{len(tweets)} tweets, approx cost ${len(tweets) * 0.00015:.2f}")