fill missing blanks in column on the attached dataset using fuzzy match

[vinayvnn]

Hi All,

I have data with three  columns, a,b, and c, based on the columns a, and b  I wanted populate data into the missing values in the column c, the method I am basically looking for is Fuzzy match, attached a sample of data however I have to replicate the same to the original data which I am working, It will be helpful If I get any leads to progress further.

a	b	c
Pension ER A Sec	Pension ER A Sec 20%	605
Period Base Sal BT	Period Base Sal BT	704
myShare Plan	myShare Plan	508
Transportation Allowance	Transportation Allowance	227
Hypo Tax (Salary)	Hypo Tax (Salary)	1001
Amount paid	Amount paid	1025
Car assistance - cash allowance	Car assistance - cash allowance	202
Hypo NI (Salary)	Hypo NI (Salary)	908
Housing Allowance	Housing Allowance	404
Pens Supp Allwnce	Pens Supp Allwnce	203
Home Allowance	Home Allowance
Assignment Allowance GRS	Assignment Allowance GRS
Sal Supp Allowance GRS	Sal Supp Allowance GRS
Accident Ins	Accident Ins
Relocation Allowance GRS	Relocation Allowance GRS
Bonus	Bonus
Total	Total
Sueldo	Sueldo
RET. TAX INCOME	RET. TAX INCOME
Plus vacacional	Plus vacacional

[gawa]

hi @vinayvnn Can I ask something?

1) I can imagine Fuzzy Match between column 'a' and 'b', but how column 'd' will be used?

2) How do you determine the value of column 'c'? Elaborate the logic, if any.

[vinayvnn]

Thanks for responding,

Basically I have a data of 5,00,000 rows in which there are 6,800 rows of blank.

Column C already have data, where as it is left with 6800 blanks this needs to be filled based on the combination of .column a and b not with d (Apologies) for the confusion.

If any of the combinations of column a and b had populated data in column c

and same combination is having blank on the 6800 rows that should be filled based on the above criteria  

[gawa]

@vinayvnn *Please try to mention so we can notice your response

If you want to find the same combination of column 'a' and 'b', and append to blank value of 'c' for that combination,

1) Split data by Filter tool; One is data having the value of 'c', another not having the value of 'c'.

2) JOIN by column 'a' and 'b' for the above two data (refer to configuration in the snapshot)

3) Combine data by Union tool

[vinayvnn]

@gawa Thanks for the quick response, however the mentioned technique have not populated all the blank values still there are few of rows with blank values, approx (4000 rows) any alternative approach would you suggest here please ? it need not to be the same as column a and b and then populate blanks on the column c.

80% of match would be acceptable