From 3dda66ce8d7ff1605f74ae41532688bc08fed974 Mon Sep 17 00:00:00 2001
From: Franck Dary <franck.dary@lis-lab.fr>
Date: Tue, 31 Mar 2020 19:54:53 +0200
Subject: [PATCH] Added script to print results
---
UD_any/print_results.py | 83 +++++++++++++++++++++++++++++++++++++++++
1 file changed, 83 insertions(+)
create mode 100755 UD_any/print_results.py
diff --git a/UD_any/print_results.py b/UD_any/print_results.py
new file mode 100755
index 0000000..44ec0af
--- /dev/null
+++ b/UD_any/print_results.py
@@ -0,0 +1,83 @@
+#! /usr/bin/python3
+
+import glob
+import sys
+import math
+
+if __name__ == "__main__" :
+
+#metrics = ["LAS","UAS","Tokens","Words","Sentences","UPOS","UFeats","Lemmas"]
+ metrics = ["Tokens","Words"]
+
+ output = []
+ outputByModelScore = dict()
+
+ for pathToFile in glob.iglob("" + '*stderr') :
+ for line in open(pathToFile, "r") :
+ if "Error" in line or "ERROR" in line or "error" in line :
+ print(pathToFile,":", file=sys.stderr)
+ print("\t"+line,end="", file=sys.stderr)
+
+ for pathToFile in glob.iglob("" + '*stdout') :
+ model = pathToFile.split("_UD_")[0]
+ corpus = pathToFile.split("_UD_")[1].split('.')[0]
+
+ if corpus not in outputByModelScore :
+ outputByModelScore[corpus] = dict()
+
+ for line in open(pathToFile, "r") :
+ for metric in metrics :
+ if metric in line and metric[0] == line[0]:
+ splited = line.strip().replace("|","").split()
+
+ model = model.split('.')[0]
+
+ if model not in outputByModelScore[corpus] :
+ outputByModelScore[corpus][model] = dict()
+ if splited[0] not in outputByModelScore[corpus][model] :
+ outputByModelScore[corpus][model][splited[0]] = []
+
+ outputByModelScore[corpus][model][splited[0]].append([corpus, splited[0], splited[3], model])
+
+ for corpus in outputByModelScore :
+ for model in outputByModelScore[corpus] :
+ for metric in outputByModelScore[corpus][model] :
+ score = 0.0
+ for exp in outputByModelScore[corpus][model][metric] :
+ score += float(exp[2])
+ score /= len(outputByModelScore[corpus][model][metric])
+ standardDeviation = 0.0
+ if len(outputByModelScore[corpus][model][metric]) > 1 :
+ for exp in outputByModelScore[corpus][model][metric] :
+ standardDeviation += (float(exp[2])-score)**2
+ standardDeviation /= len(outputByModelScore[corpus][model][metric])-1
+ standardDeviation = math.sqrt(standardDeviation)
+ if standardDeviation > 0 :
+ score = "%.2f[±%.2f]%%"%(score,standardDeviation)
+ else :
+ score = "%.2f%%"%score
+ output.append(outputByModelScore[corpus][model][metric][0])
+ output[-1][2] = score
+
+ maxColLens = [0 for _ in range(len(output[0]))]
+
+ output = [["Corpus","Metric","F1.score","Model"]] + output
+
+ for line in output :
+ for i in range(len(line)) :
+ maxColLens[i] = max(maxColLens[i], len(line[i]))
+
+ output = output[1:]
+ output.sort()
+ output = [["Corpus","Metric","F1.score","Model"]] + output
+
+ dashLine = '-' * 80
+ for i in range(len(output)) :
+ if i > 0 and output[i][0] != output[i-1][0] :
+ print(dashLine)
+ elif i > 0 and output[i][1] != output[i-1][1] :
+ print("")
+ for j in range(len(output[i])) :
+ padding = (' '*(maxColLens[j]-len(output[i][j])))+" "*3
+ print(output[i][j], end=padding)
+ print("")
--
GitLab