Open DrChandrakant opened 1 year ago
This needs to be discussed, and this issue is the perfect place so that others using and contributing to mplfinance
can see the discussion and give their opinions.
I am not fully convinced that mplfinance
should provide technical analysis, but I am also willing to consider the idea. I have some concerns, and I will list them here for consideration and discussion.
mplfinance
to provide visualization of the indicators as in the indicator examples. Also, there is at least one case that I know of (pandas-ta) where the package uses mplfinance
under the hood to provide visualization. I need to be convinced that we are providing some value-added benefit that is not necessarily provided by another half-dozen packages out there.mpf.make_addplot()
calls and whatever else is necessary to set up an indicator. I am uncertain as to whether this is benefit enough; maybe. As mentioned above, there is already at least one package that does this to some extent.
3.The following certain in my mind - the code for mpf.plot()
is already quite complicated and challenging to maintain. I am 90% certain that adding indicators to mpf.plot()
will make it extremely difficult to maintain.
mplfinance
, I would be inclined to write a new API, perhaps mpf.ta_plot()
which is a wrapper around mpf.plot()
. The user would pass in their DataFrame, and one or more indicators (names, along with adjustable parameters), and mpf.ta_plot()
would then do all the work for the caller to generate the plot. Of course users can always still call mpf.make_addplot()
and mpf.plot()
themselves directly if they choose.mplfinance
repository. Thank you! Perhaps this is the way to go, to simply continue to add and document examples of indicators here, so that others can learn and code them with their own variations and tweaks.mplfinance
depends upon. If necessary, then the new API could be a separate import
, so that users who choose not to use it can avoid the dependency. Even if we use a separate import
, I would really like to avoid adding more dependencies.mplfinance
or other visualization packages are used. Even other technical analysis packages depend on mplfinance
to do so, which demonstrates the potential of mplfinance
. this is listed under the head of the visualization I wish it stays at the same header but the subheader of mplfinance
is 'Financial Data Visualization'. Candlesticks, Bar Graph, Renko, etc are the special plot technique. But We are Missing out on Indicators properly as a visualization tool. In my view, we upgrading it for visualization purposes not for technical analysis purposes with internal calculations. Another alternative is to extend mplfinance with the indicator to go beyond the "OHLCV" data processing method so mplfinance
directly accepts data provided by another package as "OHLCV" with add column like "Ichimoke_column a, Ichimoke_column b,"mpf.plot()
and mpf.make_addplot()
accept more than 40 kwargs to modify the plot of the indicator. Even if the user asks for a feature upgrade there will be the addition of another kwargs to mpf.plot()
or mpf.make_addplot()
. which going to become difficult to maintain and adding new features getting more and more complex. For Example, if any user is willing to plot a candlestick chart with mav in the current scenario
mpf.plot( df, type="hollow_and_filled", style= 'tradingview', title = 'TCS-EQ', volume=True, volume_panel = 1, figscale =1.2, figsize = (46,26), scale_width_adjustment = {'volume' : 0.9} mav = (3,6,9), volume_alpha= (a requested feature)
upgraded version mplfinance
+ indicator concept (df+id_column, indicator = {'kind':'rsi'}) in this case mplfinance
internally determines what kind indicator to plot (RSI), where to plot determine which panel
to be used, where to fill fill_between
and what to shade. alternative make_addplot
left for those who wish to plot a custom plot.import
concern as a maintainer but a strict import
policy or template for new PR about the indicator section will element this concern. with policy or template maximum permissible import restricted to numpy and pandas which currently mplfinance is using.
mplfinance with inbuild technical parameter calculation. This Can Be a Big Achievement Because This Module is Completly Dedicated to The Finance Market Ploting.
Over The 120 Great Indicators Available Which Being Used Globally This Will be the Most Desirable Feature of mplfinance If an Indicator With Length Parameter Pass Through API Within add_plot Return a Complete Plot of the Indicator Parameter
This Feature I Have Created Locally On My system I Have Achieved 43 Primary Indicator Inclusion On My Local System Inviting Suggestions On More Inclusion. Please Go Through The Following Pull Request's Examples of New Suggestions.
Simarly Successfully integrated mplfinance Pyside6 as GUI App Sharing Notebook of Direct Indicator Inclusion API Call Through
All Example plot within mplfinance modification share notebook out but https://github.com/DrChandrakant/tr/blob/main/Future_add_on_indicator.ipynb All New Function Indicators within add_plot GUI App with mplfinance