Hello, I want to categorize about 13000 imported transactions that i have exported out of moneyplex. The import worked, but now i want to know if there is a possibility to auto or mass categorize all transactions. Havent found further information in documentation or FAQs. Using KmyMoney on Linux Debian Trixie, Version 5.1.3. Tried to find a suitable solution, but was not sucessfull.
I haven’t touched 5.1.3 in years, so if the following advice does not work it might be due to the fact that I am using 5.2.x for a long time now. You may want to switch to using the stable AppImage versions.
One possibility is to open the ledger for the account and filter the transactions by e.g. payee or other text that is common for a set of transactions. Then select them all. Not sure if Ctrl+A selects only the filtered ones. If not, then Home followed by Shift+End might do the trick. Next edit the selected transactions and only assign the category. Repeat that for each set of transactions and all accounts.
In case you have repetitive payments, you could import only a subset, do the (sometimes tedious) assignment to categories and import the other transactions after that. KMyMoney has a detection mechanism based on payees to do an auto assignment during import based on prior transactions.
In your case when you have already all transactions imported, simply delete the ones you want to re-import. You can import the full set of transactions because KMyMoney detects those that are already on file as duplicates.
Hope that helps. We’re interested about the outcome. Please feel free to report your findings.
Out of curiosity: which file format did you use to export from moneyplex?
Thank you for the quick solution. I think I understand now; it takes a few steps to assign all my transactions to categories.
-
Select the relevant transactions.
-
Manually assign the selected transactions to a suitable category.
-
“Link” the recipients to the categories in the recipient section.
-
The imported transactions, whether imported via online banking or not, will be automatically assigned in the future.
I’ve tried a few things, and it seems to work.
I installed a beta version of Moneyplex. This allowed me to use many export filters. I tried all the options: MT940, CAMT, DATEV, SML, QIF, MXL, CSV… I ultimately got the best results with the QIF format.
I tried using a newer “appimage” version of MyMoney. This caused some display errors for me.
Unfortunately, compiling the program also didn’t work. I encountered some dependency issues.
Following Error:
CMake Error at CMakeLists.txt:235 (find_package):
Could not find a configuration file for package “LibAlkimia5” that is
compatible with requested version “8.2.1”.
The following configuration files were considered but not accepted:
/usr/lib/x86_64-linux-gnu/cmake/LibAlkimia5-8.1/LibAlkimia5Config.cmake, version: 8.1.2
/lib/x86_64-linux-gnu/cmake/LibAlkimia5-8.1/LibAlkimia5Config.cmake, version: 8.1.2
Debian Trixie has an outdated version of “LibAlkimia5”.
You may also build Alkimia from source which should be straight forward and resolves your dependency problem.
Hello, thanks again for the suggested solution. I’ve spent a lot of time on Kmymoney in the last few weeks. I am of the opinion that automatic filtering by payment recipient is not a good solution. You certainly can’t expect 100% assignment of categories with the autofilter. The problem with filtering by payee is that some recipients debit different payments, for example a municipality, which charges property tax, sewage fees and perhaps also district heating fees. Filtering via note would certainly make more sense. The corresponding intended use is usually also included in this note. But this would certainly mean a lot of programming effort. Anyway, I found a bug with the automatic filter. An auto categorization is assigned to the corresponding recipient “real”. Received a booking that contained the recipient “Realschule”. The category was then not assigned, only the recipient was changed and a text was added to the information field. See also the attached screenshot. Why does the autofilter change the payee?
By the way, the attempt to compile the source code was not successful with Debian 13 trixie, the dependencies could not be fulfilled. I’m now using the app image.
Well, I agree to some extent that the current solution may not fit all use cases. Actually, what I could think of is a separation of payee matching and categorization where the latter may also consider the content of the memo field or other parts of the transaction.
I expect that matching information for the shop is not precise enough and thus also matches for the school. What does the matching tab looks like for the shop? I am pretty sure, that if you use s.th. like DANKE.*REAL would not match the school and change the payee..
For my understanding, the autofilter should only add the category of the transaction, not the payee? Am i wrong? Yes, the filter for the Real marked also fits for the scool. it was not precise enougth. anyway, the payee should not be changed. it should remain as it was transferred from the bank. it would be helpful if there would be more possibilities for auto categorysation.
I disagree. I have payees, where the bank provides different names. Also, names provided by the payee change over time. Here’s an example:
Adjusting the visible name is exactly what this feature is made for. The original is kept in the memo.
That’s a new insight for me. I thought this feature only filtered the categories. Yes, from this perspective it also makes sense to adjust the constantly changing recipient names. I have many such cases as well. What text is the recipient changed to in your example?
I use “Deutsche Post AG” but that really does not matter.