NewYour coding agent can read the release notes before it upgrades.Set up the MCP server →
PyPI · #2962 most downloaded on PyPI
Tempo is timeseries manipulation for Spark. This project builds upon the capabilities of PySpark to provide a suite of abstractions and functions that make operations on timeseries data easier and highly scalable.
Last release 1 years ago
15 Sep 2025
Release timing varies
gaps range from 9 days to 10 months
Nearly every release is documented
notes for 28 of 30 stable releases
Nothing withdrawn
no release was ever pulled
5 years old
30 releases · first in 2021
Enhanced Makefile functionality with environment checks for Python and Java setups
formatBlack command in Makefile for code formatting_cleanup_delta_warehouse method for Delta test environment cleanupOne column per quarter.
Fixed release workflows by @nfx in #416
Api-reference docs compile by @josh-melton-db in #374
Full Changelog: v0.1.26...v0.1.27
[tox] update env to avoid exit 1 error by @R7L208 in https://github.com/databrickslabs/tempo/pull/350
asofJoin() by @tnixon in https://github.com/databrickslabs/tempo/pull/363Full Changelog: https://github.com/databrickslabs/tempo/compare/v0.1.25...v0.1.26
asofJoin() by @tnixon in #363Full Changelog: v0.1.25...v0.1.26
Updating to include AS OF join fix for Delta Live Tables
Updating to include AS OF join fix for Delta Live Tables
Currently, tempo uses the underlying jvm objects to detect the size of the data frames in memory. This is not compatible with Delta Live Tables so we
Currently, tempo uses the underlying jvm objects to detect the size of the data frames in memory. This is not compatible with Delta Live Tables so we have updated the logic to skip the optimization and checks if not using the SQL optimization for small tables.
tox default environment name for type checking and flake8 configuration by @R7L208 in https://github.com/databrickslabs/tempo/pull/321Full Changelog: https://github.com/databrickslabs/tempo/compare/v0.1.23...v0.1.24
Currently, tempo uses the underlying jvm objects to detect the size of the data frames in memory. This is not compatible with Delta Live Tables so we have updated the logic to skip the optimization and checks if not using the SQL optimization for small tables.
tox default environment name for type checking and flake8 configuration by @R7L208 in #321Full Changelog: v0.1.23...v0.1.24
removed old deprecated scala code by @tnixon in #308
Full Changelog: v0.1.22...v0.1.23
Bump black from 22.8.0 to 22.10.0 by @dependabot in #262
Full Changelog: v0.1.21...v0.1.22
Bump scipy from 1.9.0 to 1.9.1 by @dependabot in #249
IntervalsDF by @R7L208 in #256Full Changelog: v0.1.20...v0.1.21
Bump pyarrow from 8.0.0 to 9.0.0 by @dependabot in https://github.com/databrickslabs/tempo/pull/238
TSDF.extractStateInterval() to perform state comparison per metric column. by @R7L208 in https://github.com/databrickslabs/tempo/pull/234Full Changelog: https://github.com/databrickslabs/tempo/compare/v0.1.19...v0.1.20
New helper methods to slice TSDFs by time, get earliest and latest records across all series
Added warning for Interpolation in long time horizon cases
Add support for milliseconds and microseconds in resample and interpolation functions
In this release, we addressed a few enhancements:
In this release, we addressed a few enhancements:
Fix to support AS OF join when sequence numbers are used. Use the proper sort order (time column, rec ind, then sequence number) as opposed to (time c
Fix to support AS OF join when sequence numbers are used. Use the proper sort order (time column, rec ind, then sequence number) as opposed to (time column, sequence number, rec ind) which results in the wrong right_event_ts.
Nothing published for this version
Add support for github pages so docs can be published formally using sphinx
Update of interpolation logic for performance (25X improvement for basic example)
Update PyPI publish to use Semver to automate version updates
Update Support for Suppressing NULL warnings in Skewed AS OF Join
Update Support for Suppressing NULL warnings in Skewed AS OF Join
Fix for multiple metric aggregation in TSDF withGroupedStats function. Docs added.
Fix for multiple metric aggregation in TSDF withGroupedStats function. Docs added.
This release addresses a recent ask for grouped statistics on a TSDF object. In particular, we know accept a frequency, group by the partition columns
This release addresses a recent ask for grouped statistics on a TSDF object. In particular, we know accept a frequency, group by the partition columns, and return the mean/sum/count/min/max/standard deviation for the grouping.
Modify code to work in Spark 3.2.1 and avoid APIs which may deprecate
Update AS OF join logic to suppress warnings and avoid collect calls with a new flag (suppress_null_warnings) for AS OF join.
Update AS OF join logic to suppress warnings and avoid collect calls with a new flag (suppress_null_warnings) for AS OF join.
Main change - Updated interpolation logic to use supplied timestamp column instead of hardcoding event_ts
Main change - Updated interpolation logic to use supplied timestamp column instead of hardcoding event_ts
Bump up PyPI Version, Update Visualizations in README
Bump up PyPI Version, Update Visualizations in README
Nothing published for this version
Resample has fill=True option for filling in zeros
added dbl-tempo to pypi (latest release v0.1.1
remove required prefix from python resample logic (specify prefix for resampling) - bars calculation uses pre-set prefixing.
Scope of enhancements
Your coding agent can read these notes before it upgrades. Set up the MCP server →