This discussion thread is to ask your questions and discuss issues, if participating in the March 2023 developer challenge for SAP Community. The topic is "Exploratory Data Analysis with SAP HANA and Python".
Please use separate discussions created for the submission of your solutions. Keep this thread for your questions only.
Hi Sergiu. Glad to see you are joining this one as well, even this time it will be super easy for you 🙂
The reason to install Client is that it comes with `hdbuserstore` used to store credentials on a client. That way participants should not think how to avoid sharing their endpoints, user names, and especially passwords, when sharing their notebooks.
Python's `hdbcli` still should be installed using pip, as then it is the latest version.
According to the introduction of the “Setup your hana database in SAP HANA Cloud trial" section, it is recommended to use the trial account, but HANA Cloud instance not support Scriptserver due to CPU/MEM resource limited under the trail account and cannot execute PAL procedures. Any suggestions on find a ML available HANA Cloud instance? Thanks in advance.
Thank you for your question. Indeed a ScriptServer is not available yet in the free BTP offering, but as well is not required for our exercises.
In this dev challenge we focus on DataFrame part of the ‘hana-ml’ package, and not on ML API. For our scope the ScriptServer is not required.
I see you added a user key on Windows machine. You can see a file where the user key is stored if you type just
But when you execute your notebook I see it tries to read the file `/root/.hdb/<name_of_machine>` from some Linux operating system using the `root` user. Are you running your notebooks in some kind of container (like Docker)?
There are might be many reasons you are getting this error on your machine: https://www.easytechjunkie.com/what-are-the-common-causes-of-socket-error-10061.htm, incl. some anti-virus programs or a network configuration.
Can you simply download the file https://raw.githubusercontent.com/SAP-samples/hana-ml-samples/main/Python-API/pal/datasets/bike.csv to your machine's directory with the notebook, and then change it to
bikes_df = pd.io.parsers.read_csv("bike.csv")
to open a CSV file from your local file system?
As well, let me attach the file with the data here as `bike.txt` to download it from this reply and rename it to `bike.csv` on your Windows machine.
Thank you for your response and providing me with the helpful suggestions for resolving the socket error that I was experiencing. I tried both of the methods that you recommended and am pleased to report that both worked successfully. I was able to save the file locally and also successfully downloaded the file you attached as bike.txt and renamed it as bike.csv on my Windows machine.
Thank you for taking the time to assist me with this issue. Your help has been greatly appreciated.