.. _cook-rfr-semi-liquid-doc:

.. ipython:: python
   :suppress:

   from rateslib import Curve, Solver, IRS, add_tenor, get_calendar, dt, get_imm
   from datetime import timedelta
   import matplotlib.pyplot as plt
   from pandas import DataFrame

Less Liquid RFR Curves (MXN, COP)
************************************

Less liquid RFR *Curve* building follows the same principles as demonstrated in
:ref:`Standard Liquid RFR Curves <cook-rfr-standard-doc>`, except data availability might be
sparser and obviously the *Instruments* must be chosen to match local conventions.

  **Key Points**

  - Standard configuration for *Curve* setup.
  - Minor modifications for the MXN market.

Example Frameworks
--------------------

.. tabs::

   .. tab:: MXN F-TIIE

      Due to data availability this example uses standard par tenor MXN F-TIIE OIS quotes
      as the *Instruuments*.

      MXN swaps have an unconventional frequency expressed in 28 calendar day periods.
      *Rateslib* handles this natively as part of its :class:`~rateslib.scheduling.Frequency`
      framework.

      .. warning::

         In practice, MXN IRS are referred to in *'months'* assuming *'one-month'* is a
         single 28-day (or 4-week) period. This gives 13 months per year and, for example a
         130-month swap would be interpreted as being the 10-year swap. *Rateslib* does
         **not** tolerate this colloquial definition and requires a more precise tenor
         to be specified. In this case *130-month* is required to be entered either
         as *"3640D"* or *"520W"*, and the ``frequency`` is set under the ``spec`` as *"28d"*
         (equivalent to *"4W"*).

      **Data**

      First, load (from source) the market data for each chosen tenor of the *Curve* and
      establish the relevant dates of construction. The below code also determines the
      appropriate maturity of each tenor under the market calendar.

      .. ipython:: python

         data = DataFrame({
             "Term": ["4W", "8W", "12W", "24W", "36W", "52W", "104W", "156W", "208W", "260W", "364W", "520W", "624W", "780W", "1040W", "1560W"],
             "Rate": [7.02, 7.017, 7.015, 6.99, 6.955, 6.965, 7.155, 7.351, 7.515, 7.645, 7.855, 8.086, 8.21, 8.355, 8.48, 8.44],
         })

         today = dt(2025, 1, 16)
         spot = get_calendar("mex").lag_bus_days(today, 2, False)
         data["Termination"] = [add_tenor(spot, _, "F", "mex") for _ in data["Term"]]

      **Curve Design**

      In order to create a perfectly solvable curve, the chosen **nodes** will be assigned
      as the maturity of each *Instrument*.

      .. ipython:: python

         ftiie = Curve(
             id="ftiie",
             convention="Act360",                          # <- important to match F-TIIE convention
             calendar="mex",                               # <- important to match F-TIIE convention
             interpolation="spline",
             nodes={
                 today: 1.0,                               # <- this is today's DF,
                 **{_: 1.0 for _ in data["Termination"][:-1]},  # <- every instrument node but the last
                 **{data["Termination"].iloc[-1] + timedelta(days=10): 1.0},  # <- avoid spline end warnings
             }
         )

      **Calibration with Solver**

      The final step puts the *Curve* and the *Instruments* together, mutating the *Curve* to match
      the *rates: s*.

      .. ipython:: python
         :okwarning:

         solver = Solver(
             curves=[ftiie],
             instruments=[IRS(spot, _, spec="mxn_irs", curves="ftiie") for _ in data["Termination"]],
             s=data["Rate"],
             instrument_labels=data["Term"],
             id="mx_rates",
         )

      Then plotting either the O/N rates or the zero rates:

      .. code-block:: python

         ftiie.plot("1b")
         ftiie.plot("Z")

      .. plot::

         from rateslib import Curve, Solver, IRS, add_tenor, get_calendar, dt
         import matplotlib.pyplot as plt
         from datetime import timedelta
         from pandas import DataFrame

         data = DataFrame({
             "Term": ["4W", "8W", "12W", "24W", "36W", "52W", "104W", "156W", "208W", "260W", "364W", "520W", "624W", "780W", "1040W", "1560W"],
             "Rate": [7.02, 7.017, 7.015, 6.99, 6.955, 6.965, 7.155, 7.351, 7.515, 7.645, 7.855, 8.086, 8.21, 8.355, 8.48, 8.44],
         })

         today = dt(2025, 1, 16)
         spot = get_calendar("mex").lag_bus_days(today, 2, False)
         data["Termination"] = [add_tenor(spot, _, "F", "mex") for _ in data["Term"]]

         ftiie = Curve(
             id="ftiie",
             convention="Act360",                          # <- important to match F-TIIE convention
             calendar="mex",                               # <- important to match F-TIIE convention
             interpolation="spline",
             nodes={
                 today: 1.0,                               # <- this is today's DF,
                 **{_: 1.0 for _ in data["Termination"][:-1]},  # <- every instrument node but the last
                 **{data["Termination"].iloc[-1] + timedelta(days=10): 1.0},  # <- avoid spline end warnings
             }
         )

         solver = Solver(
             curves=[ftiie],
             instruments=[IRS(spot, _, spec="mxn_irs", curves="ftiie") for _ in data["Termination"]],
             s=data["Rate"],
             instrument_labels=data["Term"],
             id="mx_rates",
         )

         fig1, ax1, lines = ftiie.plot("z")
         del fig1, ax1
         plt.close()
         fig, ax, _ = ftiie.plot("1b")
         ax.plot(lines[0]._x, lines[0]._y)
         plt.show()
         plt.close()