從資料來源載入資料並以 DataFrame.
語法
load(path=None, format=None, schema=None, **options)
參數
| 參數 | 類型 | 說明 |
|---|---|---|
path |
力量或列表,選擇性 | 檔案系統支援資料來源中的一條或多條路徑。 |
format |
力量,選用 | 資料來源格式。 預設為 'parquet'。 |
schema |
結構類型或力量,選用 | 輸入結構作為 StructType 物件或 DDL 格式的字串(例如 'col0 INT, col1 DOUBLE')。 |
**options |
dict | 額外的弦線選擇。 |
退貨
DataFrame
Examples
載入一個格式、結構和選項的 CSV 檔案。
import tempfile
with tempfile.TemporaryDirectory(prefix="load") as d:
df = spark.createDataFrame([{"age": 100, "name": "Alice"}])
df.write.option("header", True).mode("overwrite").format("csv").save(d)
df = spark.read.load(
d, schema=df.schema, format="csv", nullValue="Alice", header=True)
df.printSchema()
# root
# |-- age: long (nullable = true)
# |-- name: string (nullable = true)
df.show()
# +---+----+
# |age|name|
# +---+----+
# |100|NULL|
# +---+----+