How to mix and match Data

Rupak

Number City 1000 A 1100 B 1200 C 1300 D E 1400 1500 F 1300

Hi Folks,
Need your help pls.

2 Questions -

1) In the table above, you could see that 1300 is mapped to city 'D' but in last row it is blank.
How do I map it using beast mode or Magic ETL to fill by Lookup in the same column.

Req: Lookup in 'Number' and fill Blanks in 'City'

2) 5th Row contains E in Number and 1400 in City. How do I exchange it to the right position using beast mode or Magic ETL?

Req: Case when 'Number' contains String value then City.
Case when 'City' contains Numeric value then 'Number'

Thanks in Advance!

rco

First, we should solve problem #2. Then, we'll create a canonical lookup table from Number to City to solve problem #1.

Use an Add Formula tile to make two new columns, Canonical Number and Canonical City like so:

Canonical Number formula: NULLIF(STR_DIGITS(CASE WHEN City BETWEEN '0' AND '999999' THEN City ELSE Number END), '')
Canonical City formula: NULLIF(STR_REMOVE_DIGITS(CASE WHEN Number BETWEEN 'A' AND 'zzzzzz' THEN Number ELSE City END), '')

Now use a Group By tile to make the lookup table. Our one grouping column will be Canonical Number, and our one aggregate will be the First non-null value of Canonical City; we'll call this the Lookup City.

Now use the Join Data tile to Inner Join the original table (the output of the Add Formula) to the lookup table (the output of the Group By) on the Canonical Number column.

The Canonical Number column and the Lookup City column in the result of the Join Data tile have the final values you want.

The whole transform has this shape:

rco

First, we should solve problem #2. Then, we'll create a canonical lookup table from Number to City to solve problem #1.

Use an Add Formula tile to make two new columns, Canonical Number and Canonical City like so:

Canonical Number formula: NULLIF(STR_DIGITS(CASE WHEN City BETWEEN '0' AND '999999' THEN City ELSE Number END), '')
Canonical City formula: NULLIF(STR_REMOVE_DIGITS(CASE WHEN Number BETWEEN 'A' AND 'zzzzzz' THEN Number ELSE City END), '')

Now use a Group By tile to make the lookup table. Our one grouping column will be Canonical Number, and our one aggregate will be the First non-null value of Canonical City; we'll call this the Lookup City.

Now use the Join Data tile to Inner Join the original table (the output of the Add Formula) to the lookup table (the output of the Group By) on the Canonical Number column.

The Canonical Number column and the Lookup City column in the result of the Join Data tile have the final values you want.

The whole transform has this shape:

ColemenWilson

You could recreate those two fields as follows and then drag them into your card in place of the current 2 fields:

• City_2: CASE WHEN `Number` = 1300 THEN 'D' ELSE `City` END

• Number_2: CASE WHEN `Number` = 'E' THEN 1400 ELSE `Number` END

MichelleH

@Rupak I second @colemenwilson's recommendation of recreating the fields. I would do this in MagicETL using the Formula Tile. If you don't want to specify every value to swap, you could use a regular expression to match patterns based on whether `Number` contains alpha characters and `City` contains numeric characters.

City_2: CASE WHEN REGEXP_LIKE(`Number`,'[a-zA-Z].*]') then `Number` else `City` END

Number_2: CASE WHEN REGEXP_LIKE(`City`,'[0-9].*') then `City` else `Number` END