data as number

This commit is contained in:
s152483 2020-04-27 14:04:14 +00:00
parent b2176a9144
commit ff3cc25f96

View File

@ -3,28 +3,22 @@
import pandas as pd import pandas as pd
import csv import csv
import pickle import pickle
import numpy as np
def predict(): def predict():
dev0 = pd.read_csv("dev-0/in.tsv", delimiter="\t", header=None, names=["document","date"], quoting=csv.QUOTE_NONE) dev0 = pd.read_csv("dev-0/in.tsv", delimiter="\t", header=None, names=["document","date"], quoting=csv.QUOTE_NONE)["document"]
testA = pd.read_csv("test-A/in.tsv", delimiter="\t", header=None, names=["document","date"], quoting=csv.QUOTE_NONE) testA = pd.read_csv("test-A/in.tsv", delimiter="\t", header=None, names=["document","date"], quoting=csv.QUOTE_NONE)["document"]
devdoc = dev["document"]
testdoc = testA["document"]
clf = pickle.load(open("clf.model", "rb")) clf = pickle.load(open("clf.model", "rb"))
vectorizer = pickle.load(open("vectorizer.model", "rb")) vectorizer = pickle.load(open("vectorizer.model", "rb"))
dev0_vectorizer = vectorizer.transform(devdoc) dev0_vectorizer = vectorizer.transform(dev0)
testA_vectorizer = vectorizer.transform(testdoc) testA_vectorizer = vectorizer.transform(testA)
y_dev = clf.predict_proba(dev0_vectorizer) y_dev = clf.predict_proba(dev0_vectorizer)[:, 1]
y_test = clf.predict_proba(testA_vectorizer) y_test = clf.predict_proba(testA_vectorizer)[:, 1]
with open("dev-0/out.tsv", "w") as devout: np.savetxt('test-A/out.tsv', y_dev, '%f')
for line in y_dev: np.savetxt('dev-0/out.tsv', y_test, '%f')
devout.write(str(line[1])+"\n")
with open("test-A/out.tsv", "w") as testaout:
for line in y_test:
testaout.write(str(line[1])+"\n")
predict() predict()