What inspired this wish list request?
Imagine you have a list of credit card transactions from all your daily expenses and you want to recode those individual transactions into a new column of categories (Groceries, Dinning, Shopping, Public transportation...). Each of the transactions has a piece of the string that is recognizable and another piece which changes for every transaction (Walmart 637746556, Walmart 784767663, Target 6778356, FavRestaurant 736654, PizzaPlace 76354, PizzaPlace 78294...).
To use the recode platform for this task, you currently can:
- Manually group the transactions
- Use Replace String with the "Replace entire string" option to replace all the "Walmart" containing strings with "Groceries". You would need to do this for each string manually and open the Replace string window every single time.
What is the improvement you would like to see?
I would like to have a hybrid between the Replace String function and the Apply Mapping From Table function. This would consist on a table selection dialog where I can select a table that contains a column with my desired "Replace String" value I want to match and a second column with the value I want to replace the full string with.
If I had the following two check boxes in the existing Apply Mapping From Table function:
- A checkbox to tell the platform to "Find string" instead of "Matching the value"
- A checkbox to tell the platform to "Replace full string" instead of replacing the "matched substring"
That would meet the need.
Ideally, this recode would be saved as a formula in the shape of regex or something similar so when new data is added, it can auto update the recodes.
Why is this idea important?
This would be a big improvement in the speed of recoding using the recode platform for non-coders.