linerapy 0.2.1

Linear Rates Quantitative Library package for Python

linerapy

A quantitative libary for linear interest rates - LineRaPy (Pronunciation: /ˈlɪniər paɪ/). The package introduces two types related to the linear rates: YieldCurve and IRS. The YieldCurve can be constructed using the key rates, e.g. tenors and values. The IRS is constructed from the dictionary. The object of type YieldCurve can be queries for dates and zero rates. The object of type IRS can be queries for price and cash flows. The library currently supports US, USGS calendars and a vanilla IRS paying compounded SOFR vs fixed rate.

The library supports USD, EUR, GBP, SGD, JPY and CHF currencies. The corresponding generators are:

• USD SOFR SW
• GBP SONIA SW
• EUR ESTR SW
• SGD SORA SW
• JPY TONA SW
• CHF SARON SW

The valid calendars are US, USSG, EN, TG, SG, JP and CH. Here TG is the TARGET2 calendar. Supported day count conventions are

• ACT/360
• ACT/365F
• ACT/ACT
• 30/360
• 30E/360

To price the swap where the coupon already had started compounding, the user needs to provide the fixings. The user should first check the swap object for required fixing dates and provide the rate fixings for those dates.

The curve implements two interpolation methods

• LinearLogDf
• LinearRates

By default the LinearLogDf method is used.

Requirements

• Python 3.12+
• Windows / Linux

Installation

pip install linerapy

Usage

import linerapy
import datetime
import pandas as pd

if __name__ == '__main__':
    obj_dict = {
        'id': 'USD SOFR',
        'value date': datetime.date(2026, 3, 20),
        'currency': 'USD',
        'instruments': {
            'generator': ['USD SOFR SW', 'USD SOFR SW', 'USD SOFR SW', 'USD SOFR SW'],
            'tenor': ['1M', '1Y', '5Y', '10Y'],
            'quote': [3.65, 3.65, 3.65, 3.65],
            'in use': [True, True, True, True]
        }
    }

    yield_curve = linerapy.YieldCurve(obj_dict)

    print('Discount Factor:', yield_curve.df(datetime.date(2027, 3, 20)))

    df = pd.DataFrame.from_dict({
        'Date': yield_curve.dates(),
        'Zero Rate': yield_curve.rates(),
        'Discount Factor': [yield_curve.df(d) for d in yield_curve.dates()]})
    print(df)

The IRS can be created from the dictionary by providing the necessary input.

    obj_dict = {
        'id': 'IRS_001',
        'effective date': datetime.date(2026, 3, 24),
        'termination date': datetime.date(2031, 3, 24),
        'schedule roll': 'Backward',
        'leg1': {
            'currency': 'USD',
            'notional': 10000,
            'is fixed': True,
            'is paid': True,
            'rate': 0.0365,
            'calendar': 'US',
            'frequency': '6M',
            'day count convention': 'ACT/360',
            'pay lag': 2,
            'accrual adjustment': 'Modified Following',
            'accrual calendar': 'US',
            'pay adjustment': 'Modified Following',
            'pay calendar': 'US'
        },
        'leg2': {
            'currency': 'USD',
            'notional': 10000,
            'is fixed': False,
            'is paid': False,
            'index': 'SOFR',
            'margin': 0.0025,
            'calendar': 'USD',
            'frequency': '3M',
            'day count convention': 'ACT/360',
            'pay lag': 2,
            'accrual adjustment': 'Modified Following',
            'accrual calendar': 'US',
            'pay adjustment': 'Modified Following',
            'pay calendar': 'US'
        }
    }

    irs = linerapy.IRS(obj_dict)
    npv = irs.price(yield_curve)
    print(npv)

    pd.set_option('display.max_columns', None)
    pd.set_option('display.width', None)

    cf = irs.cashflows(yield_curve)

    df = pd.DataFrame.from_dict(cf['Leg1'])
    print(df)
    df = pd.DataFrame.from_dict(cf['Leg2'])
    print(df)

First request the fixing dates from the swap. Note that the valuation date is to be passed to the method.

    fixings = irs.required_fixings(datetime.date(2026, 4, 10))

Then pass the fixings to the price or cashflow method of the the swap.

    fixings_data = { 'SOFR':
                         {
                             'dates': fixings['SOFR'],
                             'values': [0.02 for d in fixings['SOFR']]
                         }
    }

    cf = irs.cashflows(yield_curve2, fixings_data)
    pvs = irs.price(yield_curve2, fixings_data)

Similarly, the first order Greeks can be calculated:

    pv01 = irs.risk(yield_curve, fixings_data)
    pv01_df = pd.DataFrame.from_dict(pv01)
    print(pv01_df)

API Reference

dates()

Returns the yield curve's dates.

Returns:

• list: List of dates.

Example:

>>> yield_curve_.dates()
[datetime.date(2026, 4, 28), datetime.date(2026, 9, 28), datetime.date(2027, 3, 26), datetime.date(2028, 3, 28)]

rates()

Returns the zero rates corresponding to the yield curve's dates.

Returns:

• list: List of zero rates.

Example:

>>> yield_curve_.rates()
[0.03694890857811573, 0.0366704955510695, 0.03634333943914385, 0.03634104468177242]

cashflows(yc, fixings)

Returns leg-wise cash flows. If the yield curve argument provided, the forward rates are estimated and the cash flows are projected.

Parameters:

• yc (YieldCurve): An optional yield curve argument.
• fixings (dict): An optional argument for fixings. If the trade has no past compounding dates, this argument is not required.

Returns:

• dict: Dictionary of leg-wise cash flows.

Example:

>>> irs.cashflows(yield_curve)
Accrual Start Date Accrual End Date Payment Date  Day Count Fraction    Rate  Notional   Cash Flow  Discount Factor  Cash Flow PV
0         2026-03-24       2026-09-24   2026-09-28            0.511111  0.0365   10000.0  186.555556         0.981029    183.016423
1         2026-09-24       2027-03-24   2027-03-26            0.502778  0.0365   10000.0  183.513889         0.963732    176.858158
2         2027-03-24       2027-09-24   2027-09-28            0.511111  0.0365   10000.0  186.555556         0.946050    176.490961
3         2027-09-24       2028-03-24   2028-03-28            0.505556  0.0365   10000.0  184.527778         0.929063    171.438002
4         2028-03-24       2028-09-25   2028-09-27            0.513889  0.0365   10000.0  187.569444         0.912291    171.117831
5         2028-09-25       2029-03-26   2029-03-28            0.505556  0.0365   10000.0  184.527778         0.895910    165.320226
6         2029-03-26       2029-09-24   2029-09-26            0.505556  0.0365   10000.0  184.527778         0.879823    162.351780
7         2029-09-24       2030-03-25   2030-03-27            0.505556  0.0365   10000.0  184.527778         0.864025    159.436635
8         2030-03-25       2030-09-24   2030-09-26            0.508333  0.0365   10000.0  185.541667         0.848426    157.418456
9         2030-09-24       2031-03-24   2031-03-26            0.502778  0.0365   10000.0  183.513889         0.833275    152.917586

price(yc, fixings)

Values the IRS using the yield curve.

Parameters:

• yc (YieldCurve): A yield curve argument.
• fixings (dict): An optional fixings argument. It is only required if the swap has past dates on which it compounds the coupon.

Returns:

• dict: Dictionary of leg-wise PV and the total NPV.

Example:

>>> npv = irs.price(yield_curve)
{'Leg1 PV': -1676.3660572375327, 'Leg2 PV': 1776.4412024912517, 'NPV': 100.07514525371903}

risk(yc, fixings)

Calculates PV01 for each key rate of the yield curve. The results are leg-wise and the netted value. The sensitivities are calculated for 1bp change. The direction of the leg, e.g. pay or receive, is taken into account when the value is return.

Parameters:

• yc (YieldCurve): A yield curve argument.
• fixings (dict): An optional fixings argument. It is only required if the swap has past dates on which it compounds the coupon.

Returns:

• dict: Dictionary of leg-wise PV01 and the Net PV01 for each key rate.

Example:

>>> pv01 = irs.risk(yield_curve, fixings_data)
  Tenor      Leg1      Leg2       Net
0    1M  0.200487  0.176671  0.377158
1    1Y  7.746666 -6.930656  0.816010
2    5Y  0.000000  0.000000  0.000000
3   10Y  0.000000  0.000000  0.000000

License

MIT License - see LICENSE for details.