Yes, just change the grouping function passed to group-by:
(group-by #(vector (% "Type") (% "Subtype")) coll)
The keys in the resulting map will be two-element vectors (e.g.
["TypeA" "SubTypeA"]) so you'll need to adjust the body of the for a
bit in BG's example.
Cheers,
Dave
On Sun, Oct 16
Thanks BG and all other responses for your help.
I have a followup question, could the same approach be extended if I
wanted to group-by multiple entries in the map? e.g. assuming the map
now had Type and Subtype and I wanted to "group-by" the pair Type and
Subtype, would the same approach work?
Thanks BG and all other responses for your help.
I have a followup question, could the same approach be extended if I
wanted to group-by multiple entries in the map? e.g. assuming the map
now had Type and Subtype and I wanted to "group-by" the pair Type and
Subtype, would the same approach work?
I would write like this
(defn sum-by-type [coll]
(->> coll
(map (fn [ {v "Value" t "Type" } ] {t (read-string v) }))
(apply (partial merge-with +
If the map keyword based, it will be simpler than above
(defn sum-by-type [coll]
(->> coll
(map (fn [ {v :value t :ty
I would write like this
(defn sum-by-type [coll]
(->> coll
(map (fn [ {v "Value" t "Type" } ] {t (read-string v) }))
(apply (partial merge-with +
If the map is keyword int map, it will be simpler than above
(defn sum-by-type [coll]
(->> coll
(map (fn [ {v :value t :type } ] {t v
Thanks for all of the useful replies, I'll try them later.
Unfortunately I don't control the input list of maps which is the
result of a web service.
On Oct 14, 10:11 pm, Sean Corfield wrote:
> On Fri, Oct 14, 2011 at 10:25 AM, der wrote:
> > Given a list of maps of the followign format: ({"Typ
On Sat, Oct 15, 2011 at 3:16 AM, Alan Malloy wrote:
> Probably CSV or some other externally-input data. And FWIW I like BG's
> solution - it's what I would have written, except it's better.
Thanks Alan :-)
Regards,
BG
--
Baishampayan Ghose
b.ghose at gmail.com
--
You received this message be
Probably CSV or some other externally-input data. And FWIW I like BG's
solution - it's what I would have written, except it's better.
On Oct 14, 2:11 pm, Sean Corfield wrote:
> On Fri, Oct 14, 2011 at 10:25 AM, der wrote:
> > Given a list of maps of the followign format: ({"Type" "A", "Value"
>
On Fri, Oct 14, 2011 at 10:25 AM, der wrote:
> Given a list of maps of the followign format: ({"Type" "A", "Value"
> "5"} {"Type" "B", "Value" "4"} {"Type" "A", "Value" "7.2"} {"Type"
> "A", "Value" "25.4"} {"Type" "B", "Value" "2.982"})
Folks are posting solutions but I wondered why your map key
> I'm a Clojure newbie trying to the following simple task:
>
> Given a list of maps of the followign format: ({"Type" "A", "Value"
> "5"} {"Type" "B", "Value" "4"} {"Type" "A", "Value" "7.2"} {"Type"
> "A", "Value" "25.4"} {"Type" "B", "Value" "2.982"})
>
> I want to compute a list of maps such th
You could first turn the Values from String into Double like this:
(defn step1 [coll]
(map (fn [m]
(assoc m "Value" (Double/parseDouble (m "Value"
coll))
Next, you can use reduce to calculate the sum per Value:
(defn step2 [coll]
(reduce (fn [res m]
(let [k (m "
Someone cleverer than me to post the built-in function that already
does exactly what you want, but here's one way:
(defn sum-by-type [in]
(->> in
(group-by #(get % "Type"))
(map
(fn [[k vs]]
{"Type" k
"Value" (reduce
(fn [acc v] (+ acc (Double/
The best i could think was:
(defn calc-types [coll]
(let [types (map (fn [m] {:type (get m "Type") :value (Double/parseDouble
(get m "Value"))}) coll) ; convert keys to keywords and values to doubles
it1 (sort-by :type types) ; sort by type
a (atom (:type (first it1)))
13 matches
Mail list logo
