I created a view 'names' over a csv file in dfs. I tried using describe statement:
describe dfs.root.`names`
Here is my create view statement:
create view dfs.root.`names` as select id,name,city from dfs.root.`names.csv`
In the output, in the data type column I get data type for all columns as 'ANY'
Here are my csv file contents:
id,name,city
1,shrinivas,pune
2,harshal,morgaon
3,nikhil,chiplun
4,ravinder,chandigarh
the extractHeader
parameter in my configuration is set to true
Any ideas on how to get the exact data type in this case?
You need to explicitly specify the datatype of each during view creation:
create view dfs.root.`names` as select cast(id as smallint) ID,cast(name as varchar) NAME,cast(city as varchar) CITY from dfs.root.`names.csv`