Sparkda tug'ilgan kuniga asoslangan holda yoshni binlar bo'yicha qanday olish mumkin?

Apache Spark'da, tug'ilgan sana bo'yicha ma'lumotlarni belgilangan sana uchun binlarga ajratish uchun quyidagi bosqichlarni bajarish mumkin:

1. Tug'ilgan sanani berilgan sana bilan ayirish orqali yoshni yillarda hisoblash.
2. Bin chegaralarini belgilash (masalan, 0-10, 11-20, 21-30 va hokazo).
3. when yoki bucketizer funksiyasidan foydalanib, yoshni binlarga taqsimlash.

PySpark misoli:

from pyspark.sql import functions as F
from pyspark.sql import SparkSession

spark = SparkSession.builder.getOrCreate()

data = [
    ("2000-05-15",),
    ("1985-10-30",),
    ("2010-01-01",)
]

df = spark.createDataFrame(data, ["birth_date"])

# Ma'lumot uchun sana
reference_date = F.lit("2024-06-01")

# Matnlarni sanaga aylantirish
df = df.withColumn("birth_date", F.to_date("birth_date"))

# Yoshni hisoblash
df = df.withColumn(
    "age",
    F.floor(F.months_between(reference_date, F.col("birth_date")) / 12)
)

# Yosh binlarini belgilash
bins = [0, 10, 20, 30, 40, 50, 60, 100]

from pyspark.ml.feature import Bucketizer

bucketizer = Bucketizer(
    splits=bins,
    inputCol="age",
    outputCol="age_bin"
)

df = bucketizer.transform(df)

df.show()

Natijada, age_bin ustunida yoshning qaysi binga tushishini ko'rsatadi. Bu indekslarni o'qilishi oson etiketlar bilan bog'lash mumkin.