I have been busy with some other projects but I managed to finish the stats analysis on my data for both 2024 Data Set and the Setup 2 Data Set ( Swing Model )
I ran the codes on Google Colab and Generated Heatmaps / Charts to Visualize my stats then I exported it to a PDF so I can access it from anywhere and take a quick look on important info before taking a trade or when im in a trade
ngl i didnt get the perfect analysis of the data ( trades i have acquired over the past to analyze in future to understand performances of my setups / models ) i did it in loops
loops mean when i exported my results to pdfs i realized i have done the same analysis for the data with just different words to call it, made me go crazy ngl since i also forget shit a lot
i did also write everything in google docs as i mentioned before in my previous postings, but it contains a lot of stuff making it hard to search or focus on something i want to look at when i want to take a trade or when im in a trade
and heres some information of wtf i mean by stats / analysis bc u would be like what is this dude talking abt he isnt even telling wtf it is
so basically lets say u collected 1000 trades that was executed / taken according to ur setup / model this is done by backtesting
then this data u logged them into a excel sheet and convert it to a .csv file
.csv file so u can run python on google colab and it can read the data in .csv file and not as a excel sheet
now doing the analysis for example is :
(i pressed "post" accidently and posted without finishing it , thats why it will show as edited in the post)
i want to know what the best days for my setup, what does that mean?
I want to know what days give me the best performance for my setup, performance = winrate (for me)
for example i can see in my data after analysing that Wednesdays have x% winrate ( x being the highest value compared to the rest of the days ) which is also the highest winrate comparing all the other weekdays , why is this useful u may ask ( or u may not ,maybe im just Scrizophrenic [idk whether i spelled it right anyways sorry abt that] after all)
well then since i know wednesday is the day with best performance i can anticipate a trade on wednesaday that will match my setup / model therefor reducing the probability of losing on other days
and it will also reduce the chance of me taking trades on less probable days so i wont lose a shit ton ( if i do consider that Wednesday is the highest probable day and if i trade on low probable days knowing Wednesday is the best to trade i will be basically gambling on other days and trading emotions which i do not recommend to myself or anyone )
just like that u can measure other metrics for example
what is the average duration of my trades
average RR of my trades ( risk : reward )
performance of my setups / models based on timings
like that, yeah first i thought this is bs is trading gambling , ig im not a expert myself either to say its not or it is , im not that good if i was i should be able to be consistently profitable without being just profitable . bc if u are gambling u wont win consistently u win sometimes u dont sometimes . so until i acheieve that i wouldnt say anything abt that
but the advantage of doing this , is having a edge or atleast SOME KNOWLEDGE about ur setup performance the market
now theres no point if u have only 2-5 weeks of data bc that doesnt cover everything , but 1 year of trading data doesnt cover everything either but gives u a atleast a lil better insight
i forgot what i was gonna say next also letting yall know i am writing according to the way thoughts comes to my mind , sometimes they wont connect when u read , sometimes they will maybe not in the same sentence , paragraph or post but sometime at some point which i mentioned at the first post of this whole thread as i remember
so yeah all of this stats just give u a better understanding of ur setup on market imo
i will have to edit atleast some of these stuff i say to make it look organized . ill try atleast with some headings or formattings
let me show a pic of a exported pdf and a example how a graph / chart should look like ( not everything just to get an understanding if someone wants to try it also )
View attachment 430254
View attachment 430257
for 2 sets of data like if u want to find best performing days or average return across days u can use a barchart to visualize like the above ones
for more than 2 sets or 4 sets like u want to find the individual performance of all setups across the weekdays u can use heatmaps like the below one
( i was gonna draw one but realized it would be not readable at all so found this on google source :
https://datasciencedojo.com/blog/heatmaps/
View attachment 430258
forgot shi again bc i was distracted when i opened that source site ended up reading it
anyways
what i tell myself is dont over complicate it bc i realized if i do that 1. i will stop doing it , 2. it would be hard for myself even when i created it myself to read it when my brain is not calm , 3. simple is better ngl
everything can be simplified if u know how to ( yea no shi sherlocks wtf was even the point in saying that fking dumbass )
and dont over do it like , bc i mentioned i was in a loop which lasted for few days bc i was doing complex analysis which was same data results but with just other words and making it myself complex so i had to write shi down in my notesbooks , same thing twice or thrice ( which i noticed after i read the book myself next day ) just to understand what i m trying to do and my goal was
and even created prims kruskals or algo charts or something to understand the procedure to myself even when im the person who made it myself ( what the fk am i saying )
i spent like 2 mins to find this image bc i forgot what to call it , its fking flow charts ( im
View attachment 430266
so yeah thats the update for now , will update along the way hope u atleast got something from the bs i said or atleast one new word or a thing
have a great weekend
.
View attachment 430267
terunoz
