Zeppelin：与字符串混合时有问题 table 显示

Question

比如我用z.show()直接输出table

%pyspark
df = spark.createDataFrame([
    (0, 0, "2018-06-03", "2018-06-03"),
    (1, 1, "2018-06-04", "2018-06-04"),
    (2, 10, "2018-06-03", None),
    (4, 1, "2018-06-05", "2018-06-01")])\
  .toDF("orderid", "customerid", "product_name", "product_name2")
print("test print string 1")
z.show(df)
z.show(df.describe())

输出很好，像这样

但是，如果我在输出 tables

之间添加一个字符串

%pyspark
df = spark.createDataFrame([
    (0, 0, "2018-06-03", "2018-06-03"),
    (1, 1, "2018-06-04", "2018-06-04"),
    (2, 10, "2018-06-03", None),
    (4, 1, "2018-06-05", "2018-06-01")])\
  .toDF("orderid", "customerid", "product_name", "product_name2")
print("test print string 1")
z.show(df)
print("test print string 2") # If I add this
z.show(df.describe())

输出变成这样，（没有显示 table）

我想知道，如何将字符串和 Zeppelin 的 table 显示一起显示？

我认为问题可能是由于字符串导致 Zeppelin 的 table 显示格式错误？

Answer 1

嗯，这是飞艇的问题。解决方法是添加“%text”以明确指定输出类型，如下所示

df = spark.createDataFrame([
    (0, 0, "2018-06-03", "2018-06-03"),
    (1, 1, "2018-06-04", "2018-06-04"),
    (2, 10, "2018-06-03", None),
    (4, 1, "2018-06-05", "2018-06-01")])\
  .toDF("orderid", "customerid", "product_name", "product_name2")
print("test print string 1")
z.show(df)
print("%text test print string 2") # If I add this
z.show(df.describe())

Zeppelin：与字符串混合时有问题 table 显示

Zeppelin: Problematic table display when mixing with strings

pyspark

apache-zeppelin