yeref 0.24.65__py3-none-any.whl → 0.24.66__py3-none-any.whl

yeref/yeref.py

@@ -16235,77 +16235,11 @@ async def return_unit_metrics(bot, PROJECT_USERNAME, EXTRA_D, BASE_P):
 
 async def return_cohort_metrics(bot, PROJECT_USERNAME, EXTRA_D, BASE_P):
     try:
-        # Получаем всех пользователей
         sql = 'SELECT USER_TID, USER_VARS, USER_LSTS FROM "USER"'
         data_users = await db_select_pg(sql, (), BASE_P)
 
-        # Собираем для каждого пользователя: месяц входа и набор месяцев активности
-        cohorts = defaultdict(set)        # cohort_month -> set(USER_TID)
-        activity_months = defaultdict(set)  # USER_TID -> set месяцов, в которые был DAU
-
-        months = ["2025-06", "2025-07", "2025-08", "2025-09"]
-        data_users = []
-        for _ in range(20):
-            # дата входа
-            entry_month = random.choice(months)
-            entry_day = random.randint(1, 28)
-            entry_date = f"{entry_month}-{entry_day:02}"
-            entry_dt_obj = datetime.strptime(entry_date, '%Y-%m-%d')
-            entry_dt = f"{entry_dt_obj.strftime('%d-%m-%Y')}_{datetime.now().strftime('%H-%M-%S')}"
-            utm = random.choice(["/start", "/startapp"])
-
-            # месяцы от входа и дальше
-            valid_months = [m for m in months if datetime.strptime(m + "-01", "%Y-%m-%d") >= entry_dt_obj.replace(day=1)]
-            if not valid_months:
-                valid_months = [entry_month]
-
-            user_mau = sorted(random.sample(valid_months, k=random.randint(1, len(valid_months))))
-            user_dau_dates = set()
-            txs, payments = [], []
-
-            # платеж
-            if user_mau:
-                pay_month = random.choice(user_mau)
-                pay_day = random.randint(1, 28)
-                pay_date = f"{pay_month}-{pay_day:02}"
-                dt_pay = datetime.strptime(pay_date, "%Y-%m-%d")
-                # if dt_pay >= entry_dt_obj:
-                payments = [{
-                    "TYPE": random.choice(["don", "sub", "pst"]),
-                    "DT_START": f"{dt_pay.strftime('%d-%m-%Y')}_14-00-00",
-                    "DT_END": "0",
-                    "AMOUNT": str(random.randint(1, 10))
-                }]
-                user_dau_dates.add(pay_date)
-
-            # вход в приложение
-            user_dau_dates.add(entry_date)
-            for m in user_mau:
-                day = random.randint(1, 28)
-                visit = f"{m}-{day:02}"
-                dt_visit = datetime.strptime(visit, "%Y-%m-%d")
-                if dt_visit >= entry_dt_obj and random.random() < 0.7:
-                    user_dau_dates.add(visit)
-
-            # статусы (отток) с низкой вероятностью
-            USER_STATUSES = []
-            if random.random() < 0.2:  # 10% шанс оттока
-                churn_month = random.choice(valid_months)
-                churn_day = random.randint(1, 28)
-                churn_date = f"{churn_month}-{churn_day:02}"
-                churn_ts = datetime.strptime(churn_date, "%Y-%m-%d").strftime("%d-%m-%Y") + "_23-59-59"
-                USER_STATUSES = [{random.choice(["left", "kicked"]): churn_ts}]
-
-            user_dau = sorted(user_dau_dates)
-            wallet = f"wallet{random.randint(1, 100)}" if txs else random.choice([f"wallet{random.randint(1, 100)}", ""])
-
-            data_users.append((
-                random.randint(100000, 999999),
-                json.dumps({"USER_WALLET": wallet, "USER_UTM": utm, "USER_DT": entry_dt}),
-                json.dumps({"USER_DAU": user_dau, "USER_MAU": user_mau, "USER_TXS": txs,
-                            "USER_PAYMENTS": payments, "USER_STATUSES": USER_STATUSES})
-            ))
-        print(f"gen {data_users=}")
+        cohorts = defaultdict(set)
+        activity_months = defaultdict(set)
 
         for USER_TID, USER_VARS, USER_LSTS in data_users:
             USER_VARS = json.loads(USER_VARS or "{}")
@@ -16313,11 +16247,12 @@ async def return_cohort_metrics(bot, PROJECT_USERNAME, EXTRA_D, BASE_P):
             USER_DT = USER_VARS.get("USER_DT", "")
             USER_DAU = USER_LSTS.get("USER_DAU", [])
 
-            # вычисляем месяц входа
+            if not USER_DT:
+                continue
+
             entry_mo = datetime.strptime(USER_DT, "%d-%m-%Y_%H-%M-%S").strftime("%Y-%m")
             cohorts[entry_mo].add(USER_TID)
 
-            # сохраняем месяцы активности
             for day_str in USER_DAU:
                 try:
                     mo = datetime.strptime(day_str, "%Y-%m-%d").strftime("%Y-%m")
@@ -16325,10 +16260,8 @@ async def return_cohort_metrics(bot, PROJECT_USERNAME, EXTRA_D, BASE_P):
                 except:
                     pass
 
-        # Список уникальных месяцев cohort, отсортированный
         cohort_months = sorted(cohorts.keys())
 
-        # Функция для прибавления месяцев
         def add_months(mo_str, n):
             y, m = map(int, mo_str.split("-"))
             total = m + n
@@ -16337,32 +16270,23 @@ async def return_cohort_metrics(bot, PROJECT_USERNAME, EXTRA_D, BASE_P):
             return f"{new_y:04d}-{new_m:02d}"
 
         num_cohorts = len(cohort_months)
-
-        # Формируем таблицу: строки M1..M{num_cohorts}
         table = []
         header = ["Месяц/Когорта"]
-        # Добавляем в заголовок каждый cohort + его размер
         for mo in cohort_months:
             header.append(f"{mo} ({len(cohorts[mo])})")
         header.append("∑")
         table.append(header)
 
-        # Для каждой строки-месяца кортаций
+        # Собираем саму таблицу
         for i in range(num_cohorts):
             row = [f"M{i+1}"]
             row_sum = 0
-            for j, cohort_mo in enumerate(cohort_months):
-                # вычисляем целевой месяц = cohort_mo + i месяцев
+            for cohort_mo in cohort_months:
                 target_mo = add_months(cohort_mo, i)
                 if i == 0:
-                    # M1 = размер когорты
                     val = len(cohorts[cohort_mo])
                 else:
-                    # считаем, сколько пользователей из когорты были активны в target_mo
-                    cnt = 0
-                    for uid in cohorts[cohort_mo]:
-                        if target_mo in activity_months.get(uid, set()):
-                            cnt += 1
+                    cnt = sum(1 for uid in cohorts[cohort_mo] if target_mo in activity_months.get(uid, set()))
                     val = cnt
                 if val:
                     row.append(str(val))
@@ -16372,14 +16296,39 @@ async def return_cohort_metrics(bot, PROJECT_USERNAME, EXTRA_D, BASE_P):
             row.append(str(row_sum))
             table.append(row)
 
+        # Считаем общий отток по всем когортам
+        total_lost = 0
+        total_start = 0
+        for cohort_mo in cohort_months:
+            counts = []
+            for i in range(num_cohorts):
+                target_mo = add_months(cohort_mo, i)
+                if i == 0:
+                    count = len(cohorts[cohort_mo])
+                else:
+                    count = sum(1 for uid in cohorts[cohort_mo] if target_mo in activity_months.get(uid, set()))
+                counts.append(count)
+            for k in range(len(counts) - 1):
+                start_cnt = counts[k]
+                next_cnt = counts[k + 1]
+                if start_cnt > 0:
+                    lost = start_cnt - next_cnt
+                    total_lost += lost
+                    total_start += start_cnt
+
+        avg_churn = (total_lost / total_start * 100) if total_start else 0
+
         path = os.path.join(EXTRA_D, "3_cohort_metrics.csv")
         with open(path, "w", newline="", encoding="utf-8") as f:
             writer = csv.writer(f)
             for r in table:
                 writer.writerow(r)
+            writer.writerow([])
+            writer.writerow([f"Avg monthly cohort churn: {avg_churn:.2f}%"])
 
         thumb = types.FSInputFile(os.path.join(EXTRA_D, "parse.jpg"))
         await bot.send_document(chat_id=my_tid, document=types.FSInputFile(path), thumbnail=thumb)
+
     except Exception as e:
         logger.info(log_ % str(e))
         await asyncio.sleep(round(random.uniform(0, 1), 2))

{yeref-0.24.65.dist-info → yeref-0.24.66.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: yeref
-Version: 0.24.65
+Version: 0.24.66
 Summary: desc-f
 Author: john smith
 Dynamic: author

yeref-0.24.66.dist-info/RECORD

@@ -0,0 +1,8 @@
+yeref/__init__.py,sha256=Qpv3o6Xa78VdLcsSRmctGtpnYE9btpAkCekgGhgJyXM,49
+yeref/l_.py,sha256=LMX_olmJwq-tgoALJCnhV_fGrL_i_43yBLkLIcEVqGo,1176743
+yeref/tonweb.js,sha256=Jf6aFOQ1OIY4q7fINYz-m5LsI3seMus124M5SYYZmtE,443659
+yeref/yeref.py,sha256=jPkWUJXO_GvUGd6SXXRnAhGNr_G0snlu5NoCgvOyI5Q,1049848
+yeref-0.24.66.dist-info/METADATA,sha256=GWW789HTQT4WP3uucPD8e7cSrrUpekFvWNjSqHXdwcg,119
+yeref-0.24.66.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+yeref-0.24.66.dist-info/top_level.txt,sha256=yCQKchWHbfV-3OuQPYRdi2loypD-nmbDJbtt3OuKKkY,6
+yeref-0.24.66.dist-info/RECORD,,

yeref-0.24.65.dist-info/RECORD

@@ -1,8 +0,0 @@
-yeref/__init__.py,sha256=Qpv3o6Xa78VdLcsSRmctGtpnYE9btpAkCekgGhgJyXM,49
-yeref/l_.py,sha256=LMX_olmJwq-tgoALJCnhV_fGrL_i_43yBLkLIcEVqGo,1176743
-yeref/tonweb.js,sha256=Jf6aFOQ1OIY4q7fINYz-m5LsI3seMus124M5SYYZmtE,443659
-yeref/yeref.py,sha256=fze5I1x5ZD0D4OgYpK1sv-bDgu2RXtMGPhDwjiwBa64,1053049
-yeref-0.24.65.dist-info/METADATA,sha256=Q5AZlJwo8T2kw62j1kSn8eFMPG3O6DBs7KoKHmLDWYo,119
-yeref-0.24.65.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-yeref-0.24.65.dist-info/top_level.txt,sha256=yCQKchWHbfV-3OuQPYRdi2loypD-nmbDJbtt3OuKKkY,6
-yeref-0.24.65.dist-info/RECORD,,