framepipes-cli 0.1.0

Pipe-friendly CLI for rolling time-window sums

framepipes-cli

Pipe-friendly CLI collection for quick data transformations on CSV streams. Small documentation tweak.

Install

pip install framepipes-cli

To install from a local checkout:

pip install -e .[dev]

Example

cat somedata.csv | fprolling 60min sum -t Time -v ValueColumn

Groupby aggregation:

cat somedata.csv | fpgroupby Category sum -v ValueColumn

Resample aggregation:

cat somedata.csv | fpresample 1D max -t Time -v ValueColumn

Filter rows with a query:

cat somedata.csv | fpquery "ValueColumn > 10"

Sort rows:

cat somedata.csv | fpsort Time

Drop columns:

cat somedata.csv | fpdrop ColA,ColB

Round float columns:

cat somedata.csv | fpround 2

Read SQL Server results:

cat query.sql | fpsqlread --server HOST --database DB --user USER --password PASS

Assign new columns:

cat somedata.csv | fpassign Total="Revenue - Discount" Ratio="Errors / Requests"

SQL Server notes:

• Requires sqlalchemy and pyodbc plus a system ODBC driver (e.g., ODBC Driver 18 for SQL Server).
• You can also pass a full SQLAlchemy URL via --url, for example: mssql+pyodbc:///?odbc_connect=DRIVER%3D%7BODBC+Driver+18+for+SQL+Server%7D%3BSERVER%3DHOST%3BDATABASE%3DDB%3BTrusted_Connection%3Dyes

Notes

• Input must be CSV with header.
• Output is CSV with the datetime column plus all aggregated numeric columns.
• Missing values in aggregated columns are filled with 0.

Sample data

See example/README.md for a runnable walkthrough using example/sample.csv.