traindiagnostics 0.3.0

A toolbox to analyse diagnostic train data!

A Python library for unevenly-spaced time series analysis in train diagnostics. Build on top of the magnificent ticts library.

Installation

pip install traindiagnostics

Want to try it out first without installing? With binder you can test out the code in an online jupyter notebook.

Usage

 import traindiagnostics as td
 ts = td.TimeSeries({
     '2019-01-01 09:00:00': 0,
     '2019-01-01 09:00:05': 1,
     '2019-01-01 09:01:02': 0,
     '2019-01-01 09:05:09': 1,
     '2019-01-01 09:05:16': 0,
     '2019-01-01 09:11:01': 1,
     '2019-01-01 09:12:59': 0,
 })

not_in_index = '2019-01-01 00:05:00'
assert ts[not_in_index] == 1  # step function, previous value

ts['2019-01-01 00:04:00'] = 10
assert ts[not_in_index] == 10

assert ts + ts == 2 * ts

ts_evenly_spaced = ts.sample(freq='1Min')

# From ticts to pandas, and the other way around
assert ts.equals(
   ts.to_dataframe().to_ticts(),
)

Contributing

Missing some features? create an issue or pull request!