Exercise: Incomplete Data |2|
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

| In a scientific calculation, matrices are commonly used for data
  representation. Suppose you are given a matrix :math:`(M\times N)`
  where you need to calculate the average of all the real numbers in the
  matrix. However, there is a problem in the matrix and some of the
  values are corrupted and given as ‘NaN’ string (stands for not a
  number).
| Write a program that skips the corrupted numbers and outputs the
  average of rest in the given matrix. 3 digits after the decimal point
  is enough for our calculation and you can use the built-in
  :math:`\color{purple}{\texttt{round()}}` function for this purpose.

-  :math:`M` and :math:`N` can be any positive integer.

-  Assume that the list of lists of floats representing the matrix will
   be given to you in variable :math:`\texttt{R}`.

*Hint:* It is necessary to keep the count of uncorrepted data in the
given matrix

.. container:: sampleio

   Sample I/O:

.. |2| image:: ../../figures/difficulty_two.png
   :class: difficulty

.. code:: default

   Input:
   R = [[0.139, 0.216, 'NaN', 0.163],
        [0.418, 0.335, 0.217, 0.419],
        [0.097, 0.252, 0.458, 0.384]]

   Output:
   0.282

   Input:
   R = [[0.855, 'NaN', 'NaN', 0.661],
        [0.713, 'NaN', 0.306, 0.805],
        ['NaN', 'NaN', 0.562, 0.1],
        ['NaN', 0.895, 0.132, 0.795],
        [0.453, 0.519, 0.662, 'NaN']]

   Output:
   0.574

   Input:
   R = [[0.464, 0.856, 'NaN', 0.126, 0.707],
        [0.822, 0.453, 0.941, 'NaN', 0.985],
        [0.535, 0.169, 'NaN', 0.334, 0.566],
        [0.136, 0.133, 0.513, 'NaN', 'NaN'],
        [0.054, 0.686, 0.19, 0.164, 0.737]]

   Output:
   0.479