Ellaborating my own trading bots to operating in the stock market and the crypto market, which are my main interest. Right now I´m developing more advanced bots in other areas such as Natural Language Processing. Meanwhile you can check my bot that trades based on volume changes and price changes in the pair BTCUSDT:
- Data has been extracted with the Yahoo! Finance package for Python called Yfinance
- For the data that is being streamed in real-time, you need to connect to Binance and obtain the permissions for the API keys.
- It includes a Blockchain data anlysis for this pair where I use to ackages pandas, matplotlib and seaborn, together with other fundamental packages such as NumPy, datetime and timedelta.
- You can go directly to the robot here. I will be preparing an updated version for short positions!
A brief introduction to the stock market operating system to give traders a spreadsheet to make some operations. It uses some uncommon packages such as the xlsxwriter to construct the Excel archives that will be sent to them. It includes:
- An equal weight distribution for the SP500 strategy to enhance performance en smaller companies
- A momentum strategy dividing them in different sets, similar to the trading bot strategy above. YOu can have a look here
- Lastly, a more conservative but solid strategy that is rooted in the fundamental analysis with KPIs such as the ROI, ROA or the total returns of the companies. This project is a bit harder than the others due to missing data and transcryption of data types for the Excel.
The company Pens And Printers had a problem with their customer focus, so I analyze the projects oriented to brainstorming.
- Data was taken from the "product_sales.csv" archive.
- It includes the whole cleaning process of the dataset, with imputation of NaN values, standarization and data quality improvement.
- For the questions that they asked me to solve, I used visualization with Seaborn and MatPlotLib.
- With pandas I performed both the Exploratory Data Analysis and the import & cleaning data processes.
- I used hypothesis tests and ANOVA tests to determine the best way to deploy offers and product sales in the company.
I have prepared this document in order to help my colleagues with the concept of hypothesis testing with some Notebooks. I already have studied this concept during my degree and I master it in my third year so I thought I could help them!
- It is divided in three notebooks that are classified as:
- Hypothesis testing fundamentals
- Proportion tests
- Non-Parametric tests
- It explains what is hypothesis testing based on the concept ELI5 from Reddit. You can check more about it on the following link
- It includes both the use in the mathematical way and in package way so it´s easily understandable from both perspectives.
- Data was taken from the feather files "dem_votes_potus_12_16.feather" and "late_shipments.feather". You can find more information about them in the notebook "hypothesis_testing_Python.ipynb".
- The main packages used are:
- Pandas
- Scipy
- Pingouin
- Statsmodels
- Other packages for imports or taking bootstrap samplings.
- The main test that you will learn in this packages are:
-
$\mathit{t}$ -test. - Two-sample
$\mathit{t}$ -test. -
$\mathit{z}$ -Proportions test. -
$\chi^{2}$ independence test. -
$\chi^{2}$ goodness of fit test. - Wilcoxon test.
- Wilcoxon-Mann-Whitney test.
- Kruskal-Wallis test.
-