A library for Time Series exploration, analysis & modelling. This includes -

As of now, this libray is in pre-alpha phase, i.e there is a lot of work still left before its first stable release.

Functionalities Include

- A mini Dashboard for Time Series Analysis, with multiple variations to each kind of analysis
- Inbuilt Freqency Variation analysis
- Intervention Analysis (In Future) 

Functionalities Include:

- Rolling Origin Framework (Currently Supports - statsmodels, sklearn, sklearn) for both multi-variate and uni-variate
- Residual Diagnostics
- Statistical Tests
- Entropy Calculations
- Intervention Analysis (In Future)

Installation

pip install tseuler

• Instantiating a DashBoard

  import pandas as pd
  import tseuler as tse
  # Read the Time Series DataFrame
  dataDF = pd.read_csv('Raw Data/stocks_data.csv', index_col=0)
  tsmadObj = tse.TSMAD(tsdata = dataDF, data_desc = 'Stocks Data',
                   target_columns = ['close'], categorical_columns = ['Name'],
                   dt_format = '%Y-%m-%d', dt_freq = 'B',
                   how_aggregate = {'open':'first', 'high':'max', 'low':'min', 'close':'last'},
                   force_interactive = True)
  tsmadObj.get_board()

tseuler has been built upon:-

• pandas
• numpy
• panel
• altair
• matplotlib
• statsmodels

v0.0.4dev0 : Development Package

• Added TSMAD
• Added TSSTATS