Class ErrorSpec

tensorflow/compiler/xla/error_spec.h:22–46

Structure describing permissible absolute and relative error bounds.

// Structure describing permissible absolute and relative error bounds.
struct ErrorSpec {
  explicit ErrorSpec(float aabs, float arel = 0, bool relaxed_nans = false)
      : abs(aabs), rel(arel), relaxed_nans(relaxed_nans) {}

  float abs;  // Absolute error bound.
  float rel;  // Relative error bound.

  // If relaxed_nans is true then any result is valid if we are expecting NaNs.
  // In effect, this allows the tested operation to produce incorrect results
  // for inputs outside its mathematical domain.
  bool relaxed_nans;

  // If this is true, then we treat each +/-inf in the actual result as
  // equivalent to our choice of either +/-inf or the min/max floating-point
  // value.
  //
  // If the expected result is +/-inf, the actual result must still be +/-inf.
  //
  // In effect, this allows the tested operation to overflow, so long as it's
  // overflowing on "large" values.
  //
  // (We could have a symmetric more_infs_ok flag if necessary; right now it
  // appears not to be.)
  bool fewer_infs_ok = false;
};

}  // namespace xla