Need halp in solving copying row value in table

Visual calculations is the way to go. 

I recreated your sample data.

For this to work you need to enable visual calculations in the options menu (preview options).

Put the fields in a matrix. press the three dots on the right upper corner and select new visual calculation: custom

You can then add in a new column calculation: 

project leader

Project leader (filled) =

IF (

    [Project] = "900"

        && [Articalname] = "fever",

    COALESCE ( PREVIOUS([project leader]), [project leader] ),

    [project leader]

)

 

Task

Task (filled) =

IF (

    [Project] = "900"

        && [Articalname] = "fever",

    COALESCE ( PREVIOUS([task]), [task] ),

    [task]

)

 

This will give you the desired outcome:

 

 

To me this feels the easiest way to accomplish your goal.

 

Hope this is helpfull.

 

Regards,

 

Chiel

 

Here is a solution that I think is following your requirements.

Here are test data I used, pulled from your snips but with an extra project to show how we resolve for when multiple projects fall on the latest previous date. Also, note that this only works if your Date column is of Date data type (or something that sorts appropriately).

Data

DAX for columns

Project Leader Filled = 
VAR _thisRank = RANK( ORDERBY( Data[Date], ASC ), PARTITIONBY( Data[Project] ) )
VAR _thisDate = Data[Date]
VAR _validRows = FILTER( 
    ALL( Data[Date], Data[Project], Data[Project Leader], Data[Task] ), 
    Data[Project Leader] <> BLANK() && Data[Task] <> BLANK() // assuming you want non-blank but remove otherwise
    && Data[ArticleName] <> "900" && Data[Date] < _thisDate 
)
VAR _validPrevious = INDEX( 
    1, _validRows, ORDERBY( 
        Data[Date], DESC,
        Data[Project], ASC // if this were DESC, then result would be "AAA"
    ) 
)
RETURN
IF(
    Data[Project] = "900" && Data[ArticleName] = "fever" && _thisRank = 1,
    CALCULATE( VALUES( Data[Project Leader] ), _validPrevious, REMOVEFILTERS( Data ) ),
    Data[Project Leader]
)

Task Filled = 
VAR _thisRank = RANK( ORDERBY( Data[Date], ASC ), PARTITIONBY( Data[Project] ) )
VAR _thisDate = Data[Date]
VAR _validRows = FILTER( 
    ALL( Data[Date], Data[Project], Data[Project Leader], Data[Task] ), 
    Data[Project Leader] <> BLANK() && Data[Task] <> BLANK() // assuming you want non-blank but remove otherwise
    && Data[ArticleName] <> "900" && Data[Date] < _thisDate 
)
VAR _validPrevious = INDEX( 
    1, _validRows, ORDERBY( 
        Data[Date], DESC,
        Data[Project], ASC // if this were DESC, then result would be "BBB"
    ) 
)
RETURN
IF(
    Data[Project] = "900" && Data[ArticleName] = "fever" && _thisRank = 1,
    CALCULATE( VALUES( Data[Task] ), _validPrevious, REMOVEFILTERS( Data ) ),
    Data[Task]
)

Output

Anonymous Thanks for your kind mail. 

I am still struggling with same problem . I needs it in as calculated column and over 4 years of data .

regards