zeta-zetra/chart_patterns — reverse-engineered prompt
Reverse engineered prompt
Build me a simple Python tool for spotting common chart patterns in trading price data. I want to load OHLC data from a CSV file with columns like open, high, low, close, then run pattern detection for double tops and bottoms, flags, head and shoulders, inverse head and shoulders, triangles including ascending, descending, and symmetrical, plus pennants.
Please make it feel like a small reusable library, not just one script. It should return the original data with the detected pattern info added, and also include an easy way to plot the results and save images when multiple matches are found. A few example scripts or usage snippets would help a lot, especially using a sample forex or crypto style dataset.
If any pattern needs a couple of tunable settings, like how strict the match should be, include that too. Add basic tests so I can trust it works. Keep setup straightforward with a requirements file or similar. If you need to check current docs or plotting best practices online, that is fine.
Want more depth? Deep Reverse