401 Client error for "multiple coins" functions

[starlordbets]

Hello,

Just wanted to let you know that there is an error being thrown for the functions involving multiple coins such as "multiple_coins_historical_data" and it gives the following error:

HTTPError: 401 Client Error: Unauthorized for url: https://api.coingecko.com/api/v3/coins/bitcoin/market_chart?vs_currency=usd&days=max&interval=daily&precision=full

If you click on that link then it gives more info on the error:

{ "error": { "status": { "timestamp": "2024-06-18T15:04:20.238+00:00", "error_code": 10012, "error_message": "Your request exceeds the allowed time range. Public API users are limited to querying historical data within the past 365 days. Upgrade to a paid plan to enjoy full historical data access: https://www.coingecko.com/en/api/pricing. " } } }

If you can update the default function parameter for "days" to be "365" instead of "max" this seems to solve the issue:

https://api.coingecko.com/api/v3/coins/bitcoin/market_chart?vs_currency=usd&days=365&interval=daily&precision=full

[nathanramoscfa]

Thank you for your interest in the pycgapi project. I am currently working on version 1.0.0 completely refactored version of pycgapi, where this issue will be fixed. If you'd like to use this version now, simply pip install the dev branch of the pycgapi GitHub repo using the pip install git+https://github.com/nathanramoscfa/pycgapi.git@dev command. I have completely refactored the project and so the project will be updated from version 0.1.4 to 1.0.0 and is not backwards compatible. Try it out, and tell me if this solves the problem you highlighted. See the examples directory for usage examples.

[starlordbets]

If I install this dev version in the same environment as the current package, will it override the previous version? I was able to tweak the package source code for the published version and got it working for my needs. Much appreciate the work you put into this. Light years better than the official Coin Gecko package.

…

On Wed, Jun 19, 2024 at 9:18 PM Nathan Ramos, CFA ***@***.***> wrote: Thank you for your interest in the pycgapi project. I am currently working on version 1.0.0 completely refactored version of pycgapi, where this issue will be fixed. If you'd like to use this version now, simply pip install the dev branch of the pycgapi GitHub repo using the pip install ***@***.*** command. I have completely refactored the project and so the project will be updated from version 0.1.4 to 1.0.0 and is not backwards compatible. Try it out, and tell me if this solves the problem you highlighted. See the examples directory for usage examples. — Reply to this email directly, view it on GitHub <#16 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/A6NDQLKGMJFZC7LB4NT3LLTZIIUWVAVCNFSM6AAAAABJQHUBZSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDCNZZGY2DOMZTGQ> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[nathanramoscfa]

@starlordbets If you're able to do what you set out to do using pycgapi 0.1.4, then great. It's stable and likely won't change, as I am focusing all my development work and improvements to pycgapi 1.0.0, yet to be released, but currently usable on the dev branch. It might not matter as you'll get the same outputs as before. pycgapi 1.0.0 is being developed more for the purpose of improving the project's maintainability and code quality, rather than adding any new features. However, CoinGecko's new ON CHAIN DEX API endpoints have been added to pycgapi 1.0.0, so if you want access to those endpoints, you'll have to use pycgapi 1.0.0 on the dev branch.

If you decided to use the dev branch on the same environment you're currently using, then yes, you'd want to override the previous version so they don't conflict. In that case, I'd recommend:

1. Backup your current environment: Use pip freeze > requirements.txt to save the current state of your environment.
2. Uninstall the current version: Run pip uninstall pycgapi.
3. Install the dev version: Use pip install git+https://github.com/nathanramoscfa/pycgapi.git@dev.

You might have to refactor your code to how pycgapi is now structured. See the examples directory in the dev for usage examples. Search the project for the endpoint you want to use, and you'll find the function for it. Let me know if you have trouble finding a particular endpoint.

Note: The source code in the dev branch is mostly finished. Working on tests and adapting the documentation to the newly refactored project. I wouldn't call it stable, but it's stable enough for you to try out and use. I tested every function in the examples directory and they all worked.

[sandeshrajbhandari]

SOLUTION
create this function in your code instead of using cg.multiple_coins_historical_data

# historical_data = cg.multiple_coins_historical_data(coin_ids, days='max')
historical_data = multiple_coins_historical_data(coin_ids, days=2)

I've just commented out days = str(days) if isinstance(days, int) or days == 'max' else 'max' since this logic is already in cg.coin_historical_market_data function. and doing this in this function turns the '10' into 'max' which is not supported for demo plans on coin gecko.
following this tutorial - https://www.coingecko.com/learn/backtesting-crypto-trading-strategies-python#install-required-libraries

from pycgapi import CoinGeckoAPI
cg = CoinGeckoAPI(coingecko_key, pro_api=pro_api)
def multiple_coins_historical_data(coin_ids: list, vs_currency: str = 'usd', days = 'max',
                                       interval: str = 'daily', precision: str = 'full', from_date: str = None,
                                       to_date: str = None, timezone_str: str = 'UTC') -> dict:
        """Get historical market data for multiple coins and separate the data into different DataFrames.

        Args:
            coin_ids (list): List of coin IDs (e.g., ['bitcoin', 'ethereum', 'tether']). Refers to coins listed in
                get_coins_list method. Required.
            vs_currency (str, optional): The target currency of market data (usd, eur, jpy, etc.). Default is 'usd'.
            days (Union[int, str], optional): Data up to a number of days ago or 'max' for maximum available data.
                Default is 'max'.
            interval (str, optional): Data interval. Possible value: 'daily'.
            precision (str, optional): Decimal precision for currency price value. Defaults to 'full'.
            from_date (str, optional): From date in 'mm-dd-yyyy' format or UNIX Timestamp. Default is None.
            to_date (str, optional): To date in 'mm-dd-yyyy' format or UNIX Timestamp. Default is None.
            timezone_str (str, optional): Timezone of the dates. Default is 'UTC'.

        Returns:
            dict: A dictionary containing DataFrames for prices, market caps, and total volumes. Each DataFrame's
                  columns represent different coins.

        Raises:
            HTTPError: If the HTTP request fails.

        Notes:
            - Cache expiration times vary: 30 seconds for 1 day, 30 minutes for 2-90 days, and 12 hours for more than
              90 days.
            - The last completed UTC day's data is available 35 minutes after midnight on the next UTC day (00:35), and
              the cache always expires at 00:40 UTC.
            - Works with the coins/{id}/market_chart and coins/{id}/market_chart/range endpoints.
            - Available for public and pro API users.

        """
        price_dfs = []
        market_cap_dfs = []
        total_volume_dfs = []

        # Rate limiting
        rate_limit = 500 if cg.pro_api else 10  # 500 req/min for Pro, 10 req/min for others
        delay = 60 / rate_limit  # Calculate delay based on rate limit

        # Handling the days parameter
        # days = str(days) if isinstance(days, int) or days == 'max' else 'max'
        # using days=10 in this function doesn't work with this.
        # removing this as this parameter is handled in cg.coin_historical_market_data() function as well

        for coin_id in tqdm(coin_ids):
            time.sleep(delay)  # Delay the request based on the rate limit
            historical_data = cg.coin_historical_market_data(coin_id, vs_currency, days, interval, precision,
                                                               from_date, to_date, timezone_str)

            # Extract each column as a separate DataFrame
            price_df = historical_data[['price']].rename(columns={'price': coin_id})
            market_cap_df = historical_data[['market_cap']].rename(columns={'market_cap': coin_id})
            total_volume_df = historical_data[['total_volume']].rename(columns={'total_volume': coin_id})

            price_dfs.append(price_df)
            market_cap_dfs.append(market_cap_df)
            total_volume_dfs.append(total_volume_df)

        # Combine the individual DataFrames
        combined_price_df = pd.concat(price_dfs, axis=1).dropna()
        combined_market_cap_df = pd.concat(market_cap_dfs, axis=1).dropna()
        combined_total_volume_df = pd.concat(total_volume_dfs, axis=1).dropna()

        return {
            'price': combined_price_df,
            'market_cap': combined_market_cap_df,
            'total_volume': combined_total_volume_df
        }