error after data import

[toubee256]

Hi there

I tried to run your script, which I think is awesome by the way.
I am getting an Error on the first Line under
## Calculate Power and Energy usable by Battery
Error on:
df = pd.DataFrame.from_records(all_data, columns = ['timestamp', 'power'])
The error is:
TypeError: Passing a DataFrame to DataFrame.from_records is not supported. Use set_index and/or drop to modify the DataFrame instead.
I get the error even when running with the provided sample data and many permutations of it (no index column, no header row, ...)

Can you give me any hint what might I might be doing wrong?
I suspected that the timestamp is of the wrong format, but I could not figure out, what it should be.

I installed Jupyter according to chri2's tutorial on a ubuntu 24.04-container in proxmox

Thanks and kind regards, Toubee

[PV-Soft]

As the error also comes with the sample data from the repo, my first guess would be that it has to do with different Pandas versions. The requirements.txt does not pin the version. Which version are you using?

[PV-Soft]

I just found out that it might be due to a wrong use of .from_records. It expects lists but not a dataframe. In earlier versions of pandas also dataframes were accepted, but recent versions throw the type error.

Try to replace
df = pd.DataFrame.from_records(all_data, columns = ['timestamp', 'power'])

with
df = all_data[['timestamp', 'power']]
or
df = all_data[['timestamp', 'power']].copy()

[toubee256]

Will try this tonight, thanks for reaching back! 24.04.2026 08:00:50 PV-Soft ***@***.***>:

…

[Bild]*PV-Soft* left a comment (PV-Soft/Battery-Simulation#5)[#5 (comment)] I just found out that it might be due to a wrong use of *.from_records*. It expects lists but not a dataframe. In earlier versions of pandas also dataframes were accepted, but recent versions throw the type error. Try to replace *df = pd.DataFrame.from_records(all_data, columns = ['timestamp', 'power'])* with *df = all_data[['timestamp', 'power']]* or *df = all_data[['timestamp', 'power']].copy()* — Reply to this email directly, view it on GitHub[#5?email_source=notifications&email_token=AX4EC5UQFJBKI2ORVWIPAKD4XL7I7A5CNFSNUABFM5UWIORPF5TWS5BNNB2WEL2JONZXKZKDN5WW2ZLOOQXTIMZRGEYDGMZWGQ42M4TFMFZW63VGMF2XI2DPOKSWK5TFNZ2KYZTPN52GK4S7MNWGSY3L#issuecomment-4311033649], or unsubscribe[https://github.com/notifications/unsubscribe-auth/AX4EC5VKJCCS3HXJ3QJRFPD4XL7I7AVCNFSM6AAAAACYEOMEAKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHM2DGMJRGAZTGNRUHE]. Triage notifications, keep track of coding agent tasks and review pull requests on the go with GitHub Mobile for iOS[https://github.com/notifications/mobile/ios/AX4EC5RZPNHCCRRGHOA2NKD4XL7I7A5CNFSNUABFM5UWIORPF5TWS5BNNB2WEL2JONZXKZKDN5WW2ZLOOQXTIMZRGEYDGMZWGQ42M4TFMFZW63VGMF2XI2DPOKSWK5TFNZ2KUZTPN52GK4S7NFXXG] and Android[https://github.com/notifications/mobile/android/AX4EC5WVM5VXT2KIG265R734XL7I7A5CNFSNUABFM5UWIORPF5TWS5BNNB2WEL2JONZXKZKDN5WW2ZLOOQXTIMZRGEYDGMZWGQ42M4TFMFZW63VGMF2XI2DPOKSWK5TFNZ2K4ZTPN52GK4S7MFXGI4TPNFSA]. Download it today! You are receiving this because you authored the thread. [Verfolgungsbild][https://github.com/notifications/beacon/AX4EC5R4FR3B2MRS2YM4R5D4XL7I7BFCNFSM6AAAAACYEOMEAKWGG33NNVSW45C7OR4XAZNMJFZXG5LFINXW23LFNZ2KUY3PNVWWK3TUL5UWJTYAAAAACAHVE4Y2M4TFMFZW63VGMF2XI2DPOI.gif]

[PV-Soft]

I've updataed the Notebook to load the data correctly and also fixed a problem with the date conversion. Let me know if it still does not work.

[toubee256]

I still get an error, this time on Line 14 of the same block:
---> 14 df['time'] = pd.to_datetime(df['timestamp'], utc=True)
with the error message:
ValueError: time data "2020-01-01T01:07:22Z" doesn't match format "%Y-%m-%dT%H:%M:%S.%f%z". You might want to try: - passing formatif your strings have a consistent format; - passingformat='ISO8601'if your strings are all ISO8601 but not necessarily in exactly the same format; - passingformat='mixed', and the format will be inferred for each element individually. You might want to use dayfirst alongside this.

I don't really understand the error message, the only difference I can see is the capital Letter "Z", which apparently should be lowercase.
However, In the provided sample_data.csv, the Letter "z" is actually lowercase.

I still am not 100% sure I did everything right when installing jupyter, since at first I got an error message about a missing library "matplotlib". With some research, I think I could resolve that issue (the error message did not reappear).

On a sidenote, I'm wondering if i could enter the timestamp-data in a different format, since I have the data in the format "YYYY-MM-DD HH:mm:ss"

Thanks and kind regards

[Phisken]

The matplot lib error should not be a problem at this point, but might be later if plotting does not work.
For now, you can try to replace the line that caused the error with one of the following options:

df['time'] = pd.to_datetime(df['timestamp'], utc=True, format='ISO8601')

or

df['time'] = pd.to_datetime(df['timestamp'], utc=True, format='mixed')

I will try pin the versions of the packages in reqirements.txt. This should eliminate the problem of inconsistent package version.

[Phisken]

I've now pinned the versions in the requirements.txt. This projects is already a few years old, so some of the versions are far away from the current versions. If you run the following code in a terminal, you should be able to create a python environment with the versions that I used (thanks to @chri2):

python3 -m venv venv
. venv/bin/activate
pip3 install -r requirements.txt

In a quick test the install command failed because of an old version of pip, which you could update with:

pip install --upgrade pip

Afterwards, run the install command again.

Now you can start the notebook with:

jupyter-notebook