--------------------------------------------
lines 5-59 of file: example/user/compress.py
--------------------------------------------

# {xrst_begin user_compress.py}
# {xrst_spell
#     def
# }
# {xrst_comment_ch #}
#
# Using Data Interval Compression
# ###############################
#
# Purpose
# *******
# This example shows how to use the
# :ref:`compression intervals<option_table@compress_interval>` option.
#
# Integrands
# **********
# For this example there are one integrand, ``Sincidence`` .
#
# Nodes
# *****
# There is only one node called ``world`` for this example.
# There are no random effects because there are no child nodes.
#
# True Iota
# *********
# For this example, the true model incidence rate *iota* is
# {xrst_code py}
def iota_true(age) :
    return 0.01 * ( 1 + ((age - 50) / 50)**2 )
# {xrst_code}
#
# Model
# *****
# There is only one rate *iota* and it piecewise linear in age
# with knots at the age points 0, 50, and 100.
#
# Data
# ****
# There is one data point measuring Sincidence
# with the true value of *iota* at times 0.0, 50, 100.
# The corresponding age intervals are [0,0], [0,100], [100,100].
# The age interval for the second measurement should be [50,50]
# and using interval compression with make it so.
# This is a cooked up example where interval compression makes the solution
# more accurate. Under normal circumstances, the answer less accurate
# but faster to compute.
#
# Source Code
# ***********
# {xrst_literal
#     BEGIN PYTHON
#     END PYTHON
# }
#
# {xrst_end user_compress.py}