Prepare a Jupyter Notebook script

Super Protocol TEE file system

Inside the TEE, you have a special file structure:

Location	Purpose	Access
/sp/inputs/input-0001	Possible data location 1	Read-only
/sp/inputs/input-0002	Possible data location 2	Read-only
/sp/output	Output folder for your results	Read own files and write

important

Always use absolute paths, such as /sp/....

1. Prepare your script

In the Jupyter Notebook:

1. Locate your input file in /sp/inputs.

2. Load the data; for example, from a CSV file.

3. Process the data as needed.

4. Write the results to /sp/output so they are saved and returned after execution.

2. Find the input file

In the TEE, locate your input file, for example input.csv, using the find_input_file() function. This function checks both input folders and returns the absolute path:

from pathlib import Path

# Helper function to find the input file 
def find_input_file(filename):
locations = [
Path(f"/sp/inputs/input-0002/{filename}"),
Path(f"/sp/inputs/input-0001/{filename}")
]
for path in locations:
if path.exists():
print(f"Using input file: {path}")
return path
raise FileNotFoundError(f"{filename} not found in input-0002 or input-0001")

# Specify the input file (replace "input.csv" with the name of your file) 
input_path = find_input_file("input.csv")

3. Read the input file

You can load the input file using any library or method suitable for your data format. For example, with pandas:

import pandas as pd

df = pd.read_csv(input_path)
print(df.head())

Use input_path as the file location if you use another method, such as the csv module, a JSON parser, or any other library.

4. Process and save the data

After loading your data, process it in memory and save the results to /sp/output. Ensure your processed data, whether a DataFrame, list, dictionary, or other format, is ready before saving. For example:

from pathlib import Path
import pandas as pd

# Create the output directory if it doesn't exist
output_path = Path("/sp/output")
output_path.mkdir(parents=True, exist_ok=True)
# Specify the output file (replace "output.csv" with the name of your file)
output_file = output_path / "output.csv"
df.to_csv(output_file, index=False)