Datasets:
File size: 3,873 Bytes
6d08226 |
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 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 |
import datetime
from enum import Enum
from pathlib import Path
import bs4
from typing import Generator, Iterable, Optional
from dataclasses import dataclass
from .stocks import StockMarketAnalyzer, load_stock_prices
def clean_paragraph(paragraph: str):
return paragraph.strip().replace("\n", " ").replace("\r", "").replace("--", "")
def process_transcript(transcript: str, min_words: int = 30):
paragraphs = transcript.split("\n\n")
for i, paragraph in enumerate(paragraphs):
if len(paragraph.split()) > min_words:
yield i, clean_paragraph(paragraph)
class Sentiment(str, Enum):
positive = "positive"
negative = "negative"
def __repr__(self) -> str:
return self.value
def __str__(self) -> str:
return self.value
@dataclass
class EarningsPrompt:
company: str
date: datetime.datetime
para_no: int
label: Sentiment
text: str
def to_dict(self):
return {
"company": self.company,
"date": self.date,
"para_no": self.para_no,
"label": self.label,
"text": self.text,
}
@dataclass
class EarningsCall:
company: str
date: datetime.datetime
file_path: Path
sentiment: Optional[Sentiment] = None
transcript: Optional[str] = None
def load_transcript(self):
if self.transcript is None:
self.transcript = bs4.BeautifulSoup(
self.file_path.read_text(), "html.parser"
).text
def generate_prompts(self, min_words: int = 30):
if self.sentiment is None:
raise ValueError("EarningsCall sentiment must be set")
if self.transcript is None:
self.load_transcript()
for para_no, paragraph in process_transcript(
self.transcript, min_words=min_words
):
yield EarningsPrompt(
company=self.company,
date=self.date,
para_no=para_no,
label=self.sentiment,
text=paragraph,
)
def set_sentiment(
self,
market_analyzer: StockMarketAnalyzer,
days_before: int = 1,
days_after: int = 1,
):
beats_market = market_analyzer.beats_market(
self.company, self.date, days_before=days_before, days_after=days_after
)
self.sentiment = Sentiment.positive if beats_market else Sentiment.negative
@classmethod
def from_file(cls, path: Path):
"""
Given a path to an earnings call transcript file, extracts the company name, date,
and file path and returns an EarningsCall object containing this information.
Args:
path (Path): The path to the earnings call transcript file.
Returns:
EarningsCall: An object containing the company name, date, and file path.
"""
company = path.parent.stem
date = datetime.datetime.strptime(path.stem[: -(1 + len(company))], "%Y-%b-%d")
return cls(company=company, date=date, file_path=path)
def load_earnings_calls(files: Iterable[Path]) -> Generator[EarningsCall, None, None]:
return (EarningsCall.from_file(f) for f in files)
def process_earnings_calls(
market_data_directory: Path, days_before: int = 1, days_after: int = 1
) -> Generator[EarningsCall, None, None]:
stock_prices = load_stock_prices(market_data_directory.glob("**/*.csv"))
earnings_calls = load_earnings_calls(market_data_directory.glob("**/*.txt"))
market_analyzer = StockMarketAnalyzer(stock_prices)
for call in earnings_calls:
call.set_sentiment(market_analyzer, days_before, days_after)
yield call
def generate_prompts(calls: Iterable[EarningsCall], min_words: int = 30):
for call in calls:
yield from call.generate_prompts(min_words=min_words)
|