rdm | Relational data mining in python

I finally found what I was doing wrong. I was confusing file "ownership" and file "permissions".

setfacl is used to set default "permissions" for files created in a directory. What I actually needed was to set default "ownership". This is done by setting the "setgid flag" with the chmod command after properly setting the directory group and user ownership.

I wanted all newly created files in my project directory to belong to the user "john" and the group "www-data".

You have code to switch from 1 to 0, but no code to switch from 0 to 1.
Below code alternates between 1 and 0 with each run of the loop.
I also changed to jrepl (more modern and powerful). It isn't necessary (though possible) to process piped data and redirect the result to another file. The /f switch gives the inputfile to process, the /o switch gives the outputfile. By giving it a single -, it uses the same filename as the input file (and overwrites it with the new(changed) data).

Despite the API is done and even documented, the implementation is missing. Take a look. 8-byte SIMD vectors is essential part of NEON ISA for decades now (was introduced in 2005), yet the .NET runtime only implements them when compiling for ARM64 (released in 2013).

I don’t work for Microsoft and have no idea how exactly they compile their binaries, but the source code tells they have at least some support for NEON when building for ARM64 target. If you want these intrinsics in .NET, you can try the 64-bit OS.

There’s a workaround — implement your performance-critical pieces in C++, compile a shared library for Linux, then use [DllImport] to consume these functions from .NET. I have built non-trivial Linux software that way (example), using the following gcc flags to build the DLLs: -march=native -mfpu=neon-fp16 -mfp16-format=ieee -ffast-math -O3 -fPIC This way it will work for 32-bit OS, and won’t require anything special from .NET runtime, I’ve tested with .NET Core 2.1.

The implication is that if only one session is returned, $response.userId and $response.sessionName are not one-element arrays but just strings - and indexing into strings returns characters (e.g., 'foo'[0] returns 'f').^[1]

You can use @(), the array-subexpression operator to ensure that you're always dealing with arrays:

Well it turns out, trying to replicate the issue on a smaller scale made my brain neurons fire enough, that i figured out the problem. My model column count can change and it does change, however, I had not written anything that notifies about column count changing beginRemoveColumns and endRemoveColumns and beginInsertColumns and endInsertColumns. I implemented those in my code like so

Try modifying Form tag’s attributes in html

action=“BE URL or api Path” method=“POST”

Reference: https://www.w3schools.com/tags/att_form_method.asp

There is an easier and more efficient approach ...

When you want to call a python function from another file, you should not try to run another python process by calling os.system. Just import that function: from plotrdm import plot_rdm

Instead of using os.filewalk and a file.startswith check, we can cleanup the code a lot by using the nice python library glob. I throw in a enumerate for good measure.

Your new rdmsave.py

You can use a capturing group based regex like