`os.chdir` fails when script invoked as `python x` rather than `./x` about pi-eco-indicator HOT 11 CLOSED

To get the error I just cd to the dir and type python store_data.py
I cant show any output as all I get is the error.

I SSH from Mac terminal as well if that changes anything.

from pi-eco-indicator.

Ah, I was only running through python to export (>>) to the text file.
However, ive just tried it again and now get

Working in Octopus Agile import mode.
Agile version set: £1 cap, new-new formula (current)
Selected region G
Connected to database...
API request successful, status 200.
100 prices were inserted, ending at 22:00 on Saturday 17 Jun.
1 unneeded data points from the past were deleted.

As said earlier, it does seem to be working now so not sure why the graph messed up previously.

I assume that even with this change, the crontab script will still auto run the software when needed?

from pi-eco-indicator.

Ill try and get the time to have a look, always difficult looking at someone elses code but ill certainly see what I can do :)

from pi-eco-indicator.

Hello. It sounds very much like your installation isn't updating its database, for whatever reason.

Please could you show me the output of ./store_data.py and crontab -l as well as the contents of eco_indicator.log?

Please surround any text you paste in backticks ` so that it formats clearly :)

from pi-eco-indicator.

I left it going all day today and then at some point the graph returned and has been there ever since with no issues.

crontab -l gives me:

@reboot /bin/sleep 40; /usr/bin/python3 /home/pi/pi-eco-indicator/update_display.py > /home/pi/pi-eco-indicator/eco_indicator.log 2>&1
*/30 * * * * /bin/sleep 5; /usr/bin/python3 /home/pi/pi-eco-indicator/update_display.py > /home/pi/pi-eco-indicator/eco_indicator.log 2>&1
57 16 * * * /usr/bin/python3 /home/pi/pi-eco-indicator/store_data.py > /home/pi/pi-eco-indicator/eco_indicator.log 2>&1
57 18 * * * /usr/bin/python3 /home/pi/pi-eco-indicator/store_data.py > /home/pi/pi-eco-indicator/eco_indicator.log 2>&1
57 20 * * * /usr/bin/python3 /home/pi/pi-eco-indicator/store_data.py > /home/pi/pi-eco-indicator/eco_indicator.log 2>&1

Also just tried to run store_data.py manually to export output to a text file and I got the following in terminal:

  File "/home/pi/pi-eco-indicator/store_data.py", line 215, in <module>
    os.chdir(os.path.dirname(sys.argv[0]))
FileNotFoundError: [Errno 2] No such file or directory: ''

However when I reboot the pi, it auto runs and runs fine :)
Hope that all helps in some way

from pi-eco-indicator.

You need the backtick, the one to the left of the number 1 key, below Esc :) I've edited your post to fix that.

OK so we definitely have an error when running store_data.py...

Can you show me the entire sequence - the command you run as well as the output?

from pi-eco-indicator.

Ah, I see! I (and the instructions) invoke it via ./store_data.py as the script is marked as executable.

However it is entirely legit to invoke it as python store_data.py.

Can you edit store_data.py for me, changing line 215 from

os.chdir(os.path.dirname(sys.argv[0]))

to

os.chdir(sys.path[0])

and see if that fixes your issue? :)

from pi-eco-indicator.

Hmm, why do you need to export to a text file? Is that a useful feature for you?

Did you try it again with or without python?

Did you have a go at implementing my change? I'm testing it here to check it doesn't mess with the cron job, but I don't see any reason why it would.

When it is playing silly buggers, eco_indicator.log should be helpful.

The graph didn't mess up, it ran out of data, because store_data.py was failing. If the time at the top of the screen is correct (well, correct for the last 30 mins) then update_display.py is working fine. It's designed to be tolerant of missing data so it'll just display less and less until it runs out. Perhaps I should add a message on there that says "oops, no data!"

(triple backticks on the lines above and below your paste snippet will preserve the line breaks!)

from pi-eco-indicator.

Hmm, why do you need to export to a text file? Is that a useful feature for you?
Its just the way I have always grabbed output!

Tried it again with or without python?
Tried it with both and seems fine

Did you have a go at implementing my change? I'm testing it here to check it doesn't mess with the cron job, but I don't see any reason why it would.
Yes, and so far all seems great, the pi is resetting every 30 mins

The graph didn't mess up, it ran out of data, because store_data.py was failing. If the time at the top of the screen is correct (well, correct for the last 30 mins) then update_display.py is working fine. It's designed to be tolerant of missing data so it'll just display less and less until it runs out. Perhaps I should add a message on there that says "oops, no data!"

Possibly a good idea that as will stop it looking 'broken' and rather like its trying but not finding the data it needs.

(triple backticks on the lines above and below your paste snippet will preserve the line breaks!)
As you have probably guessed, quite new to all this GitHub stuff, but thankful for people like yourself who post on here with some great programs :)

from pi-eco-indicator.

Interesting, I just copy and paste straight from the terminal! Sounds like we have a fix for at least a partial issue, so I'll implement that now... do you fancy having a go at a "no data" message? 🤔

from pi-eco-indicator.

Fixed in 5335f1d

from pi-eco-indicator.