File size: 3,242 Bytes
ba77815 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 |
import pandas as pd
import ast
import numpy as np
import os
from pathlib import Path
import argparse
# path="/home/rjalota/climabench_data/CARDS2_multisource_multilabel_data.csv"
# path = "/home/rjalota/climabench_data/data/training"
claim_mapping = {
"0": "No claim",
"1": "Global warming is not happening",
"2": "Human greenhouse gases are not causing climate change",
"3": "Climate impacts/global warming is beneficial/not bad",
"4": "Climate solutions won’t work",
"5": "Climate movement/science is unreliable"
}
subclaim_mapping = {
"0_0": "No claim",
"1_1": "Ice/permafrost/snow cover isn’t melting",
"1_2": "We’re heading into an ice age/global cooling",
"1_3": "Weather is cold/snowing",
"1_4": "Climate hasn’t warmed/changed over the last (few) decade(s)",
"1_6": "Sea level rise is exaggerated/not accelerating",
"1_7": "Extreme weather isn’t increasing/has happened before/isn’t linked to climate change",
"2_1": "It’s natural cycles/variation",
"2_3": "There’s no evidence for greenhouse effect/carbon dioxide driving climate change",
"3_1": "Climate sensitivity is low/negative feedbacks reduce warming",
"3_2": "Species/plants/reefs aren’t showing climate impacts/are benefiting from climate change",
"3_3": "CO2 is beneficial/not a pollutant",
"4_1": "Climate policies (mitigation or adaptation) are harmful",
"4_2": "Climate policies are ineffective/flawed",
"4_4": "Clean energy technology/biofuels won’t work",
"4_5": "People need energy (e.g. from fossil fuels/nuclear)",
"5_1": "Climate-related science is unreliable/uncertain/unsound (data, methods & models)",
"5_2": "Climate movement is unreliable/alarmist/corrupt"
}
def parse_args():
parser = argparse.ArgumentParser(description='run binary classifer')
parser.add_argument("--path", default="data/training", help="path to exeter training dir. containing train, test, validation splits")
parser.add_argument("--out", default="climaEval/exeter/", help="output directory path")
return parser.parse_args()
if __name__ == '__main__':
args = parse_args()
for filename in os.listdir(args.path):
print(filename)
f = os.path.join(args.path, filename)
df = pd.read_csv(f, header=0)
# print(df.head())
df.rename(columns={"claim": "sub_claim_code"}, inplace=True)
df["claim_code"] = df["sub_claim_code"].str.split("_").str[0]
df["claim"] = df["claim_code"].map(claim_mapping)
df["sub_claim"] = df["sub_claim_code"].map(subclaim_mapping)
claim_df = df[['text', 'claim_code', 'claim']]
subclaim_df = df[['text', 'sub_claim_code', 'sub_claim']]
claim_df = claim_df.dropna()
Path(f"{args.out}/claim/").mkdir(parents=True, exist_ok=True)
Path(f"{args.out}/sub_claim/").mkdir(parents=True, exist_ok=True)
claim_df.to_csv(f"{args.out}/claim/{filename}", index=False)
subclaim_df.to_csv(f"{args.out}/sub_claim/{filename}", index=False)
#print(claim_df.claim.value_counts())
print("---")
print(claim_df.claim_code.value_counts())
print(sorted(list(subclaim_df["sub_claim_code"].unique()))) |