Scraping the Recommendation Rating dates

[tamireiny]

Hello, great toolbox!
I wanted to raise need to get the dates analysts set price target and recommendation, currently the recommendation mean is available via the API, but might be very outdated at certain cases, by adding dates of the latest upgrades or downgrade it will be like confidence measure for the mean value.

ticker_info = yf.Ticker(ticker)
ticker_info_data = ticker_info.info
yahoo_rating = ticker_info_data['recommendationMean']

Attached images to help demonstrate:

Thanks for your attention :)

[bot-unit]

I don't know about data of price targets, but recommendation rating was added #1754
Upgrades and downgrades is planned for release 0.3, but we can add early if @ValueRaider permits.

                       firm        toGrade     fromGrade action
epochGradeDate                                                 
2023-11-01          Wedbush     Outperform    Outperform   reit
2023-10-31          Needham            Buy           Buy   reit
2023-10-30          Keybanc  Sector Weight                 reit
2023-10-30         Barclays   Equal-Weight  Equal-Weight   main
2023-10-30      Oppenheimer     Outperform    Outperform   main

[tamireiny]

@bot-unit Thanks for quick response. Can you please add a code snippet that show how to reach the table you mentioned?

[ValueRaider]

@bot-unit Release 0.3 can wait

Edit: nonsense? 0.3 should be independent from dev -> main -> release

Edit2: I think I get question. Ideally feature stays branch until get some feedback from eager users @tamireiny

[bot-unit]

@tamireiny
for current dev branch (not for release)
Add exceptions, errors checking and other things

import yfinance as yf
import pandas as pd


def parse_upgrades_downgrades(data: dict) -> pd.DataFrame:
    data = data.get('history', None)
    if data is None:
        raise ValueError('history not found in data')
    df = pd.DataFrame(data)
    df.set_index('epochGradeDate', inplace=True)
    df.index = pd.to_datetime(df.index, unit='s')
    return df


if __name__ == '__main__':
    ticker = yf.Ticker("AAPL")
    data = ticker._quote._fetch(proxy=None, modules=["upgradeDowngradeHistory"])
    top_level = data.get('quoteSummary')
    if top_level is None:
        raise ValueError('quoteSummary not found in data')
    result = top_level['result'][0]
    df = parse_upgrades_downgrades(result['upgradeDowngradeHistory'])
    print(df.head())

[tamireiny]

Thanks @bot-unit , I am using the latest release, so I understand that it's not available the moment.
Any idea when this part will be merged to main?

[bot-unit]

I don't known. I can write snippet for current release. It's will be not clear code but will helps you now until release 0.3

[ValueRaider]

Read #1080

[bot-unit]

@tamireiny

you should understand that is bad and dangerous code, but python permit doing it.
i will create pull request in few days

from types import MethodType
from yfinance import Ticker
import pandas as pd


def fetch_upgrades_downgrades(self):
    params_dict = {"modules": "upgradeDowngradeHistory", "corsDomain": "finance.yahoo.com", "symbol": self._symbol}
    result = self._data.get_raw_json("https://query2.finance.yahoo.com/v10/finance/quoteSummary/", user_agent_headers=self._data.user_agent_headers, params=params_dict, proxy=None)
    return result


def get_upgrades_downgrades(self) -> pd.DataFrame:
    resp = self._quote.fetch_upgrades_downgrades()
    top_level = resp.get('quoteSummary')
    if top_level is None:
        raise ValueError('quoteSummary not found in data')
    try:
        data = top_level['result'][0]['upgradeDowngradeHistory']['history']
        df = pd.DataFrame(data)
        df.set_index('epochGradeDate', inplace=True)
        df.index = pd.to_datetime(df.index, unit='s')
        return df
    except KeyError:
        raise ValueError('upgradeDowngradeHistory not found in data')


def build_ticker(ticker: str) -> Ticker:
    t = Ticker(ticker)
    t._quote.fetch_upgrades_downgrades = MethodType(fetch_upgrades_downgrades, t._quote)
    t.get_upgrades_downgrades = MethodType(get_upgrades_downgrades, t)
    return t


if __name__ == "__main__":
    ticker = build_ticker("AAPL")
    print(ticker.get_upgrades_downgrades())

[tamireiny]

@bot-unit Works well, you are awesome! 👑
Do we know how if yahoo finance takes recommendation date into account in the recommendation mean? or it's simply: 1 strong buy, 2 buy, 3 hold, 4 sell and 5 strong sell ?

[bot-unit]

@tamireiny
Let's try to visit tipranks or refinitiv

[bot-unit]

@tamireiny done at #1773

[tamireiny]

@bot-unit Thanks for update!