csv_series_libraryThe CSV Series Library is an innovative tool designed for Pine Script developers to efficiently parse and handle CSV data for series generation. This library seamlessly integrates with TradingView, enabling the storage and manipulation of large CSV datasets across multiple Pine Script libraries. It's optimized for performance and scalability, ensuring smooth operation even with extensive data.
Features:
Multi-library Support: Allows for distribution of large CSV datasets across several libraries, ensuring efficient data management and retrieval.
Dynamic CSV Parsing: Provides robust Python scripts for reading, formatting, and partitioning CSV data, tailored specifically for Pine Script requirements.
Extensive Data Handling: Supports parsing CSV strings into Pine Script-readable series, facilitating complex financial data analysis.
Automated Function Generation: Automatically wraps CSV blocks into distinct Pine Script functions, streamlining the process of integrating CSV data into Pine Script logic.
Usage:
Ideal for traders and developers who require extensive data analysis capabilities within Pine Script, especially when dealing with large datasets that need to be partitioned into manageable blocks. The library includes a set of predefined functions for parsing CSV data into usable series, making it indispensable for advanced trading strategy development.
Example Implementation:
CSV data is transformed into Pine Script series using generated functions.
Multiple CSV blocks can be managed and parsed, allowing for flexible data series creation.
The library includes comprehensive examples demonstrating the conversion of standard CSV files into functional Pine Script code.
To effectively utilize the CSV Series Library in Pine Script, it is imperative to initially generate the correct data format using the accompanying Python program. Here is a detailed explanation of the necessary steps:
1. Preparing the CSV Data:
The Python script provided with the CSV Series Library is designed to handle CSV files that strictly contain no-space, comma-separated single values. It is crucial that your CSV file adheres to this format to ensure compatibility and correctness of the data processing.
2. Using the Python Program to Generate Data:
Once your CSV file is prepared, you need to use the Python program to convert this file into a format that Pine Script can interpret. The Python script performs several key functions:
Reads the CSV file, ensuring that it matches the required format of no-space, comma-separated values.
Formats the data into blocks, where each block is a string of data that does not exceed a specified character limit (default is 4,000 characters). This helps manage large datasets by breaking them down into manageable chunks.
Wraps these blocks into Pine Script functions, each block being encapsulated in its own function to maintain organization and ease of access.
3. Generating and Managing Multiple Libraries:
If the data from your CSV file exceeds the Pine Script or platform limits (e.g., too many characters for a single script), the Python script can split this data into multiple blocks across several files.
4. Creating a Pine Script Library:
After generating the formatted data blocks, you must create a Pine Script library where these blocks are integrated. Each block of data is contained within its function, like my_csv_0(), my_csv_1(), etc. The full_csv() function in Pine Script then dynamically loads and concatenates these blocks to reconstruct the full data series.
5. Exporting the full_csv() Function:
Once your Pine Script library is set up with all the CSV data blocks and the full_csv() function, you export this function from the library. This exported function can then be used in your actual trading projects. It allows Pine Script to access and utilize the entire dataset as if it were a single, continuous series, despite potentially being segmented across multiple library files.
6. Reconstructing the Full Series Using vec :
When your dataset is particularly large, necessitating division into multiple parts, the vec type is instrumental in managing this complexity. Here’s how you can effectively reconstruct and utilize your segmented data:
Definition of vec Type: The vec type in Pine Script is specifically designed to hold a dataset as an array of floats, allowing you to manage chunks of CSV data efficiently.
Creating an Array of vec Instances: Once you have your data split into multiple blocks and each block is wrapped into its own function within Pine Script libraries, you will need to construct an array of vec instances. Each instance corresponds to a segment of your complete dataset.
Using array.from(): To create this array, you utilize the array.from() function in Pine Script. This function takes multiple arguments, each being a vec instance that encapsulates a data block. Here’s a generic example:
vec series_vector = array.from(vec.new(data_block_1), vec.new(data_block_2), ..., vec.new(data_block_n))
In this example, data_block_1, data_block_2, ..., data_block_n represent the different segments of your dataset, each returned from their respective functions like my_csv_0(), my_csv_1(), etc.
Accessing and Utilizing the Data: Once you have your vec array set up, you can access and manipulate the full series through Pine Script functions designed to handle such structures. You can traverse through each vec instance, processing or analyzing the data as required by your trading strategy.
This approach allows Pine Script users to handle very large datasets that exceed single-script limits by segmenting them and then methodically reconstructing the dataset for comprehensive analysis. The vec structure ensures that even with segmentation, the data can be accessed and utilized as if it were contiguous, thus enabling powerful and flexible data manipulation within Pine Script.
Library "csv_series_library"
A library for parsing and handling CSV data to generate series in Pine Script. Generally you will store the csv strings generated from the python code in libraries. It is set up so you can have multiple libraries to store large chunks of data. Just export the full_csv() function for use with this library.
method csv_parse(data)
Namespace types: array
Parameters:
data (array)
method make_series(series_container, start_index)
Namespace types: array
Parameters:
series_container (array)
start_index (int)
Returns: A tuple containing the current value of the series and a boolean indicating if the data is valid.
method make_series(series_vector, start_index)
Namespace types: array
Parameters:
series_vector (array)
start_index (int)
Returns: A tuple containing the current value of the series and a boolean indicating if the data is valid.
vec
A type that holds a dataset as an array of float arrays.
Fields:
data_set (array) : A chunk of csv data. (A float array)
Portfolio management
TradingPortfolioLibrary "TradingPortfolio"
Simple functions for portfolio management. A portfolio is essentially
a float array with 3 positions that gets passed around
into these functions that ensure it gets properly updated as trading ensues.
An example usage:
import hugodanielcom/TradingPortfolio/XXXX as portfolio
var float my_portfolio = portfolio.init(0.0, strategy.initial_capital) // Initialize the portfolio with the strategy capital
if close < 10.0
portfolio.buy(my_portfolio, 10.0, close) // Buy when the close is below 10.0
plot(portfolio.total(my_portfolio), title = "Total portfolio value")
get_balance(portfolio) Gets the number of tokens and fiat available in the supplied portfolio.
Parameters:
portfolio : A portfolio float array as created by the `init()` function.
Returns: The tokens and fiat in a tuple
set_balance(portfolio, new_crypto, new_fiat) Sets the portfolio number of tokens and fiat amounts. This function overrides the current values in the portfolio and sets the provided ones as the new portfolio.
Parameters:
portfolio : A portfolio float array as created by the `init()` function.
new_crypto : The new amount of tokens in the portfolio.
new_fiat : The new amount of fiat in the portfolio
Returns: The tokens and fiat in a tuple
init(crypto, fiat) This function returns a clean portfolio. Start by calling this function and pass its return value as an argument to the other functions in this library.
Parameters:
crypto : The initial amount of tokens in the portfolio (defaults to 0.0).
fiat : The initial amount of fiat in the portfolio (defaults to 0.0).
Returns: The portfolio (a float )
crypto(portfolio) Gets the number of tokens in the portfolio
Parameters:
portfolio : A portfolio float array as created by the `init()` function.
Returns: The amount of tokens in the portfolio
fiat(portfolio) Gets the fiat in the portfolio
Parameters:
portfolio : A portfolio float array as created by the `init()` function.
Returns: The amount of fiat in the portfolio
retained(portfolio) Gets the amount of reatined fiat in the portfolio. Retained fiat is not considered as part of the balance when buying/selling, but it is considered as part of the total of the portfolio.
Parameters:
portfolio : A portfolio float array as created by the `init()` function.
Returns: The amount of retained fiat in the portfolio
retain(portfolio, fiat_to_retain) Sets the amount of fiat to retain. It removes the amount from the current fiat in the portfolio and marks it as retained.
Parameters:
portfolio : A portfolio float array as created by the `init()` function.
fiat_to_retain : The amount of fiat to remove and mark as retained.
Returns: void
total(portfolio, token_value) Calculates the total fiat value of the portfolio. It multiplies the amount of tokens by the supplied value and adds to the result the current fiat and retained amount.
Parameters:
portfolio : A portfolio float array as created by the `init()` function.
token_value : The fiat value of a unit (1) of token
Returns: A float that corresponds to the total fiat value of the portfolio (retained amount included)
ratio(portfolio, token_value) Calculates the ratio of tokens / fiat. The retained amount of fiat is not considered, only the active fiat being considered for trading.
Parameters:
portfolio : A portfolio float array as created by the `init()` function.
token_value : The fiat value of a unit (1) of token
Returns: A float between 1.0 and 0.0 that corresponds to the portfolio ratio of token / fiat (i.e. 0.6 corresponds to a portfolio whose value is made by 60% tokens and 40% fiat)
can_buy(portfolio, amount, token_value) Asserts that there is enough balance to buy the requested amount of tokens.
Parameters:
portfolio : A portfolio float array as created by the `init()` function.
amount : The amount of tokens to assert that can be bought
token_value : The fiat value of a unit (1) of token
Returns: A boolean value, true if there is capacity to buy the amount of tokens provided.
can_sell(portfolio, amount) Asserts that there is enough token balance to sell the requested amount of tokens.
Parameters:
portfolio : A portfolio float array as created by the `init()` function.
amount : The amount of tokens to assert that can be sold
Returns: A boolean value, true if there is capacity to sold the amount of tokens provided.
buy(portfolio, amount, token_value) Adjusts the portfolio state to perform the equivalent of a buy operation (as in, buy the requested amount of tokens at the provided value and set the portfolio accordingly).
Parameters:
portfolio : A portfolio float array as created by the `init()` function.
amount : The amount of tokens to buy
token_value : The fiat value of a unit (1) of token
Returns: A boolean value, true the requested amount of tokens was "bought" and the portfolio updated. False if nothing was changed.
sell(portfolio, amount, token_value) Adjusts the portfolio state to perform the equivalent of a sell operation (as in, sell the requested amount of tokens at the provided value and set the portfolio accordingly).
Parameters:
portfolio : A portfolio float array as created by the `init()` function.
amount : The amount of tokens to sell
token_value : The fiat value of a unit (1) of token
Returns: A boolean value, true the requested amount of tokens was "sold" and the portfolio updated. False if nothing was changed.
