Nrthugu

I'm trying to run a python script. It runs a trained model from keras.
I worked with a C# Web API. I've cracked my head trying to make the python script run. I've made it through a .bat file. It works perfectly on a windows form project. And the bat itself works fine as well.
The issue comes when the Web API wants to run the .bat file, for some reason
it when I run the script through IIS Express it won't recognize the data.csv file that the model is supposed to read.

Let me show you some code
Run the bat file

           System.Diagnostics.Process.Start(@"pathtobatfilerunner.bat");

bat file

pathtoenviromentpython.exe pathtoscriptAI.py %*

PAUSE

Allright, so when I run it outside IIS Express it works fine but when I run it on the Web API ...

Traceback (most recent call last):
File "D:AIASSETSAI.py", line 13, in
q = pd.read_csv('data.csv')
File "C:UsersMarcoAnaconda3envsentornotensorlibsite-packagespandasioparsers.py", line 678, in parser_f
return _read(filepath_or_buffer, kwds)
File "C:UsersMarcoAnaconda3envsentornotensorlibsite-packagespandasioparsers.py", line 440, in _read
parser = TextFileReader(filepath_or_buffer, **kwds)
File "C:UsersMarcoAnaconda3envsentornotensorlibsite-packagespandasioparsers.py", line 787, in init
self._make_engine(self.engine)
File "C:UsersMarcoAnaconda3envsentornotensorlibsite-packagespandasioparsers.py", line 1014, in _make_engine
self._engine = CParserWrapper(self.f, **self.options)
File "C:UsersMarcoAnaconda3envsentornotensorlibsite-packagespandasioparsers.py", line 1708, in init
self._reader = parsers.TextReader(src, **kwds)
File "pandas_libsparsers.pyx", line 384, in pandas._libs.parsers.TextReader.cinit
File "pandas_libsparsers.pyx", line 695, in pandas._libs.parsers.TextReader._setup_parser_source
FileNotFoundError: File b'data.csv' does not exist

but it does actually exist, this just happens when I run it from the Web API.

some extra information: the bat file takes a long time to run the python script. but the script itself runs instantaneously.
the data.csv file is generated through code and it generates correctly. Could it be that the file takes too long to write itself and the batch file runs before the data.csv is written? I made some tests in another project but the data.csv was written instantaneously.

Thanks in advance
Edit:
I've just checked debugging and even with a Thread sleep that the data.csv is not taking too long to build. IIS console just won't let the python script read the data.csv file

Have you tried changing the account the website uses in IIS? As it could be a permission issue if its using the networkservice account.

I found out what the issue was.
as I'm running the script outside the environment, and I'm calling the interpreter from a Process class, there's no way for the script to know what is the working folder. It assumes it to be where the interpreter is. the solution is very straight forward, just add the path in the script.
something like this:

q = pd.read_csv('D:\AIASSETS\data.csv')

and wherever you use a file add "D:AIASSETS" or better said whatever your working folder is.
even in your model or if you're writing a file.
I'm not to worded on python, but I believe there's a way to set the working folder by code as well, that might work too.