Hi Rashmi,

Here is one way to handle invalid values in the date column

rawData = Import["~/Downloads/sample.csv"]

(* Add column headers and convert to Dataset *)
dataset = 
 AssociationThread[{"Id", "Name", "Date", "Subject", "MessageId"} -> #] & /@ rawData // Dataset

(* Helper function to check if a string is convertible to DateObject *)
checkDate[date_] := 
 If[Quiet@Check[DateObject[date], False] === False, False, DateObject[date]]

(* Convert Date column to DateObject where possible, filter out unconverted *)
cleanDates = 
 dataset[All, <|#, "Date" -> checkDate[#Date]|> &][Select[#Date =!= False &]]

Other filter examples

(* Only rows starting with "compat:" (ignoring case) in the Subject column *)
compatData = 
 cleanDates[Select[StringStartsQ[#Subject, "compat:", IgnoreCase -> True] &]]

(* Name is sophia *)
sophiaData = cleanDates[Select[#Name == "sophia" &]]

(* With "version" <anything> "999" in the Subject *)
v999Data = 
 cleanDates[Select[StringContainsQ[#Subject, ___ ~~ "version" ~~ ___ ~~ "999"] &]]

You can also use regular expressions for string matching and filtering.

Hi Eric, the solution provided by Rohit helped out. My challenge was to first check if the column is even convertible for the format i want, for example here, whether date column can be converted into date object. I am able to filter out the csv data. I was able to complete the task. Thank you once again for all you kind help. It is very very appreciated.

Hello Rohit, Thank you for your quick answer. I have attached a sample copy of the data. This is a CSV file in which the date column is having different format for a lot of the records, i want to remove the rows where the date is not in the format that can be converted to date object. The problem is not the date format, if I were to perform another operation on a different column where I am required to delete rows which doesn't have the desired form of value, how do I go about it in a dataset without converting it into a database. For example, for any column that contains special characters, or contains space, etc. Thanks.

Attachments:

sample.csv

Oh, wait. Is this a variable update question? Do you want the variable cleanfinal to be updated with the results of your query? That's easy:

cleanfinal = cleanfinal[Select[StringLength[#Date] == 19 &]]

Hello Eric,

Thank you for your quick answer. But my problem is not this particular query. If i were to perform any Select query for example:

cleanfinal[Select[#Date === "String" &], All]

This gives me the result. But i cannot figure out a way to perform deletecases on the dataset. Do i need to convert the dataset into a database anytime i were to perform Queries similar to SQL or any relational database?