Metadata-Version: 2.1
Name: traces
Version: 0.6.3
Summary: A Python library for unevenly-spaced time series analysis
Home-page: https://github.com/stringertheory/traces
Author: Mike Stringer
Author-email: mike.stringer.internet@gmail.com
Requires-Python: >=3.8,<4.0
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.8
Classifier: Programming Language :: Python :: 3.9
Classifier: Programming Language :: Python :: 3.10
Classifier: Programming Language :: Python :: 3.11
Classifier: Programming Language :: Python :: 3.12
Requires-Dist: infinity (>=1.5,<2.0)
Requires-Dist: sortedcontainers (>=2.4.0,<3.0.0)
Project-URL: Documentation, https://stringertheory.github.io/traces/
Project-URL: Repository, https://github.com/stringertheory/traces
Description-Content-Type: text/markdown

<!-- [![Version](https://img.shields.io/pypi/v/traces.svg?)](https://pypi.python.org/pypi/traces) -->
<!-- [![PyVersions](https://img.shields.io/pypi/pyversions/traces.svg)](https://pypi.python.org/pypi/traces) -->
<!-- [![Documentation Status](https://readthedocs.org/projects/traces/badge/?version=master)](https://traces.readthedocs.io/en/master/?badge=master) -->
<!-- [![Release](https://img.shields.io/github/v/release/stringertheory/traces)](https://img.shields.io/github/v/release/stringertheory/traces) -->

[![Build status](https://img.shields.io/github/actions/workflow/status/stringertheory/traces/main.yml?branch=main)](https://github.com/stringertheory/traces/actions/workflows/main.yml?query=branch%3Amain)
[![codecov](https://codecov.io/gh/stringertheory/traces/branch/main/graph/badge.svg)](https://codecov.io/gh/stringertheory/traces)
[![Commit activity](https://img.shields.io/github/commit-activity/y/stringertheory/traces)](https://img.shields.io/github/commit-activity/m/stringertheory/traces)

# traces

A Python library for unevenly-spaced time series analysis.

## Why?

Taking measurements at irregular intervals is common, but most tools are
primarily designed for evenly-spaced measurements. Also, in the real
world, time series have missing observations or you may have multiple
series with different frequencies: it can be useful to model these as
unevenly-spaced.

Traces was designed by the team at
[Datascope](<[https://datascopeanalytics.com/](https://en.wikipedia.org/wiki/Datascope_Analytics)>) based on several practical
applications in different domains, because it turns out [unevenly-spaced
data is actually pretty great, particularly for sensor data
analysis](https://traces.readthedocs.io/).

## Installation

To install traces, run this command in your terminal:

```shell
$ pip install traces
```

## Quickstart: using traces

To see a basic use of traces, let's look at these data from a light
switch, also known as _Big Data from the Internet of Things_.

![](docs/_static/img/trace.svg)

The main object in traces is a [TimeSeries](https://traces.readthedocs.io/en/master/api_reference.html#timeseries), which you
create just like a dictionary, adding the five measurements at 6:00am,
7:45:56am, etc.

```pycon
>>> time_series = traces.TimeSeries()
>>> time_series[datetime(2042, 2, 1,  6,  0,  0)] = 0 #  6:00:00am
>>> time_series[datetime(2042, 2, 1,  7, 45, 56)] = 1 #  7:45:56am
>>> time_series[datetime(2042, 2, 1,  8, 51, 42)] = 0 #  8:51:42am
>>> time_series[datetime(2042, 2, 1, 12,  3, 56)] = 1 # 12:03:56am
>>> time_series[datetime(2042, 2, 1, 12,  7, 13)] = 0 # 12:07:13am
```

What if you want to know if the light was on at 11am? Unlike a python
dictionary, you can look up the value at any time even if it's not one
of the measurement times.

```pycon
>>> time_series[datetime(2042, 2, 1, 11,  0, 0)] # 11:00am
0
```

The `distribution` function gives you the fraction of time that the
`TimeSeries` is in each state.

```pycon
>>> time_series.distribution(
>>>   start=datetime(2042, 2, 1,  6,  0,  0), # 6:00am
>>>   end=datetime(2042, 2, 1,  13,  0,  0)   # 1:00pm
>>> )
Histogram({0: 0.8355952380952381, 1: 0.16440476190476191})
```

The light was on about 16% of the time between 6am and 1pm.

### Adding more data...

Now let's get a little more complicated and look at the sensor readings
from forty lights in a house.

![](docs/_static/img/traces.svg)

How many lights are on throughout the day? The merge function takes the
forty individual `TimeSeries` and efficiently merges them into one
`TimeSeries` where the each value is a list of all lights.

```pycon
>>> trace_list = [... list of forty traces.TimeSeries ...]
>>> count = traces.TimeSeries.merge(trace_list, operation=sum)
```

We also applied a `sum` operation to the list of states to get the
`TimeSeries` of the number of lights that are on.

![](docs/_static/img/count.svg)

How many lights are on in the building on average during business hours,
from 8am to 6pm?

```pycon
>>> histogram = count.distribution(
>>>   start=datetime(2042, 2, 1,  8,  0,  0),   # 8:00am
>>>   end=datetime(2042, 2, 1,  12 + 6,  0,  0) # 6:00pm
>>> )
>>> histogram.median()
17
```

The `distribution` function returns a [Histogram](https://traces.readthedocs.io/en/master/api_reference.html#histogram) that
can be used to get summary metrics such as the mean or quantiles.

### It's flexible

The measurements points (keys) in a `TimeSeries` can be in any units as
long as they can be ordered. The values can be anything.

For example, you can use a `TimeSeries` to keep track the contents of a
grocery basket by the number of minutes within a shopping trip.

```pycon
>>> time_series = traces.TimeSeries()
>>> time_series[1.2] = {'broccoli'}
>>> time_series[1.7] = {'broccoli', 'apple'}
>>> time_series[2.2] = {'apple'}          # puts broccoli back
>>> time_series[3.5] = {'apple', 'beets'} # mmm, beets
```

## More info

To learn more, check the [examples](https://traces.readthedocs.io/en/master/examples.html) and the detailed [reference](https://traces.readthedocs.io/en/master/api_reference.html#).

## Contributing

Contributions are welcome and greatly appreciated! Please visit our [guidelines](https://github.com/datascopeanalytics/traces/blob/master/CONTRIBUTING.md)
for more info.

