README improvements

Manual.ipynb

@@ -73,7 +73,7 @@
    "source": [
     "# Flat view\n",
     "\n",
-    "FProfile's `flat` report is a [dataframe](http://juliadata.github.io/DataFrames.jl/stable/man/getting_started/#Getting-Started-1), however no particular knowledge of dataframes is necessary; I'll provide a few common operations below."
+    "FProfile's `flat` report is a [dataframe](http://juliadata.github.io/DataFrames.jl/stable/man/getting_started/#Getting-Started-1), however no particular knowledge of dataframes is necessary. I'll provide a few common operations below."
    ]
   },
   {
@@ -161,7 +161,7 @@
     "    df[df[:function].===derivative, :]\n",
     "```\n",
     "\n",
-    "It is common to focus optimization efforts on one module at a time (... the one you're developing!) `flat(pd, my_module)` filters out other modules and adds a useful column: `end_count_percent` measures how much `my_module`-specific work is done on that line.\n",
+    "It is common to focus optimization efforts on one module at a time (... the one you're developing). `flat(pd, MyModule)` filters out other modules and adds a useful column: `end_count_percent` measures how much `MyModule`-specific work is done on that line.\n",
     "\n",
     "For instance, in the code below, while the `do_computation()` call takes a long time (it has a high `count_percent`), it merely calls another function, so it has a low `end_count_percent`. `sum_of_sin` has `end_count_percent = ~87%` because while it calls `sum` and `sin`, those are defined in another module (`Base`), and counted as external.\n",
     "\n",

README.md

@@ -6,15 +6,17 @@
 
 [![codecov.io](http://codecov.io/github/cstjean/FProfile.jl/coverage.svg?branch=master)](http://codecov.io/github/cstjean/FProfile.jl?branch=master)
 
-FProfile.jl is a nicer interface for Julia's [sampling profiler](https://docs.julialang.org/en/latest/manual/profile/). 
+FProfile.jl is a nicer interface for Julia's [sampling
+profiler](https://docs.julialang.org/en/latest/manual/profile/). It defines a few
+profiling metrics not available in `Base.Profile`'s reports.
 
 # Manual
 
 [User manual](http://nbviewer.jupyter.org/github/cstjean/FProfile.jl/blob/master/Manual.ipynb)
 
 # Related efforts
 
-- [ProfileView.jl](https://github.com/timholy/ProfileView.jl) is really good, especially
-  if your code is hierarchical.
+- [ProfileView.jl](https://github.com/timholy/ProfileView.jl) is a really good way to
+  visualize profiling data, especially if your code is hierarchical.
 - [TraceCalls.jl](http://nbviewer.jupyter.org/github/cstjean/TraceCalls.jl/blob/master/README.ipynb#Profiling) can be useful to track down memory allocations and type-stability issues.