Fix some spacing/formatting issues

Author: timholy

blog/_posts/2016-03-27-arrays-iteration.md

@@ -70,8 +70,7 @@ vectorized functions" strategy common to other languages. What we lack
 is a "sparse iteration API" that lets you write the main algorithms of
 sparse linear algebra efficiently in a generic way.  Our current model
 is probably fine for `SparseLike*Dense` operations, but gets to be
-harder to manage if you want to efficiently compute, e.g., `Bidiagonal
-* SparseMatrixCSC`: the number of possible combinations you have to
+harder to manage if you want to efficiently compute, e.g., `Bidiagonal*SparseMatrixCSC`: the number of possible combinations you have to
 support grows rapidly with more sparse types, and thus represents a
 powerful incentive for developing efficient, generic low-level
 operations.
@@ -130,8 +129,7 @@ It's also worth pointing out some problems:
 
 - Most importantly, it requires that one adopt a slightly different
   programming style. Despite being well into another release cycle,
-  this transition is still [not complete, even in Base]
-  (https://github.com/JuliaLang/julia/pull/15434#issuecomment-194991739).
+  this transition is still [not complete, even in Base](https://github.com/JuliaLang/julia/pull/15434#issuecomment-194991739).
 
 - For algorithms that involve two or more arrays, there's a
   possibility that their "best" iterators will be of different