diff --git a/ranking/generate_csv.py b/ranking/generate_csv.py
new file mode 100644
index 0000000..050107d
--- /dev/null
+++ b/ranking/generate_csv.py
@@ -0,0 +1,60 @@
+from collections import defaultdict
+import json
+import os
+
+import numpy as np
+
+
+def generate_from_folders(base_path, mode="sample", avg_cases=True):
+
+    sub_fldrs = [el for el in sorted(os.listdir(base_path), reverse=True) if os.path.isdir(os.path.join(base_path, el))]
+
+    submission_dict_brain = dict()
+    submission_dict_colon = dict()
+
+    for fldr in sub_fldrs:
+        fldr_path = os.path.join(base_path, fldr)
+
+        if (
+            not os.path.exists(os.path.join(fldr_path, f"brain_{mode}_scores.json"))
+            or not os.path.exists(os.path.join(fldr_path, f"colon_{mode}_scores.json"))
+            or not os.path.exists(os.path.join(fldr_path, f"config.json"))
+        ):
+            continue
+
+        with open(os.path.join(fldr_path, "config.json"), "r") as fp_:
+            username = json.load(fp_)["username"]
+
+        if username in submission_dict_brain:
+            continue
+
+        with open(os.path.join(fldr_path, f"brain_{mode}_scores.json"), "r") as fp_:
+            brain_scores = json.load(fp_)
+        with open(os.path.join(fldr_path, f"colon_{mode}_scores.json"), "r") as fp_:
+            colon_scores = json.load(fp_)
+
+        if avg_cases:
+            brain_scores = [np.mean(brain_scores)]
+            colon_scores = [np.mean(colon_scores)]
+
+        submission_dict_brain[username] = brain_scores
+        submission_dict_colon[username] = colon_scores
+
+        print(username, fldr)
+
+    ranking_csv = os.path.join(base_path, f"{mode}_ranking.csv")
+    if not avg_cases:
+        ranking_csv = os.path.join(base_path, f"{mode}_ranking_noavg.csv")
+    with open(ranking_csv, "w") as fp_:
+        fp_.write('"task","alg_name","value","case"\n')
+
+    with open(ranking_csv, "a") as fp_:
+        for u_name, subs in submission_dict_brain.items():
+            for i, val in enumerate(subs):
+                fp_.write(f'"brain","{u_name}",{val},"{i}"\n')
+        for u_name, subs in submission_dict_colon.items():
+            for i, val in enumerate(subs):
+                fp_.write(f'"colon","{u_name}",{val},"{i}"\n')
+
+    print("Done")
+
diff --git a/ranking/generate_ranking.r b/ranking/generate_ranking.r
new file mode 100644
index 0000000..e193074
--- /dev/null
+++ b/ranking/generate_ranking.r
@@ -0,0 +1,40 @@
+library(challengeR)
+
+##### Sample
+
+data=read.csv("sample_ranking.csv")
+challenge=as.challenge(data,by="task",
+                        algorithm="alg_name", case="case", value="value",smallBetter = FALSE
+                        )
+ranking=challenge%>%aggregateThenRank(FUN = mean, 
+                                      na.treat=0,
+                                      ties.method = "min" 
+)
+meanRanks=ranking%>%consensus(method = "euclidean")
+ranking%>%report(consensus=meanRanks,
+         title="MOOD 2020 - Sample",
+         file = "sample_report",
+         format = "PDF", # format can be "PDF", "HTML" or "Word"
+         latex_engine="pdflatex"#LaTeX engine for producing PDF output. Options are "pdflatex", "lualatex", and "xelatex"
+         )
+
+
+
+##### Pixel
+
+data=read.csv("pixel_ranking.csv")
+challenge=as.challenge(data,by="task",
+                        algorithm="alg_name", case="case", value="value",smallBetter = FALSE
+                        )
+ranking=challenge%>%aggregateThenRank(FUN = mean, 
+                                      na.treat=0,
+                                      ties.method = "min" 
+)
+meanRanks=ranking%>%consensus(method = "euclidean")
+ranking%>%report(consensus=meanRanks,
+         title="MOOD 2020 - Pixel",
+         file = "pixel_report",
+         format = "PDF", # format can be "PDF", "HTML" or "Word"
+         latex_engine="pdflatex"#LaTeX engine for producing PDF output. Options are "pdflatex", "lualatex", and "xelatex"
+         )
+