diff --git a/.gitignore b/.gitignore
new file mode 100644
index 00000000000..ae05230ef93
--- /dev/null
+++ b/.gitignore
@@ -0,0 +1,3 @@
+
+.DS_Store
+.Rhistory
diff --git a/cachematrix.R b/cachematrix.R
index a50be65aa44..f84a8900b28 100644
--- a/cachematrix.R
+++ b/cachematrix.R
@@ -1,15 +1,47 @@
 ## Put comments here that give an overall description of what your
 ## functions do
 
-## Write a short comment describing this function
+## makeCacheMatrix creates an object, which keeps both
+## original matrix and inverse one. 
 
 makeCacheMatrix <- function(x = matrix()) {
-
+    inv <- NULL ## variable to keep inverse matrix. 
+    ## set original matrix. Note: it does not calculate the inverse matrix
+    ## by default.
+    set <- function(y) { 
+        x <<- y
+        inv <<- NULL
+    }
+    get <- function() x ## get original matrix
+    setinv <- function(inverse) inv <<- inverse ## set inverse matrix
+    getinv <- function() inv ## get inverse matrix
+    list(set = set, get = get,
+         setinv = setinv,
+         getinv = getinv)
 }
 
 
-## Write a short comment describing this function
+## cacheSolve returns inverse matrix for x, which is 'cachematrix'. 
+## if there is no cached inverse matrix, it calculates new one.
 
 cacheSolve <- function(x, ...) {
-        ## Return a matrix that is the inverse of 'x'
+    ## Return a matrix that is the inverse of 'x'
+    
+    m <- x$getinv()
+    
+    ## check if there is the previously set value, 
+    ## because calculating inverse matrix might be expensive.
+    if(!is.null(m)) { 
+        message("getting cached data")
+        return(m)
+    }
+    
+    m <- x$get()
+    
+    ## calculate inverse matrix only if the original one is not null
+    if (!is.null(m)) {
+        m <- solve(m, ...) ## reusing m, so we can return its value
+        x$setinv(m)
+    }
+    m
 }