finx

A native desktop and WebAssembly financial data visualization tool built on Dear ImGui and ImPlot. Load data from CSV files, HTTP APIs, or Yahoo Finance, combine streams with formula expressions, and explore them interactively in resizable plot windows.

Features

• Multiple data sources — CSV upload, HTTP GET (JSON/CSV), Yahoo Finance tickers, and formula-derived streams
• Interactive plots — line, scatter, bar, and step chart types with dual Y-axes and time-scaled X-axis
• Formula engine — arithmetic expressions with scalar and rolling window functions over any loaded streams
• Crosshair tooltip — hover over any plot to see exact Y values for each series at the nearest data point
• Data table view — right-click any stream to inspect raw data in a scrollable table with virtual rendering
• CSV export — export any loaded or derived stream to a CSV file (native builds)
• Dockable UI — ImGui docking layout with persistent window positions via imgui.ini
• Persistent config — streams, plots, and field mappings saved to finx.json between sessions
• Web build — same codebase compiles to WebAssembly via Emscripten

Prerequisites

macOS

brew install sdl2
# libcurl is provided by the system

Linux

sudo apt install build-essential libsdl2-dev libcurl4-openssl-dev

Web (Emscripten)

# Install and activate the Emscripten SDK — https://emscripten.org/docs/getting_started/downloads.html
source /path/to/emsdk/emsdk_env.sh

Yahoo Finance (optional)

pip install pybind11 yfinance

When pybind11 is importable at build time the Makefile automatically enables the integration.

Build

Download the single-header dependencies (only needed once):

make deps

Native desktop app

make          # build → build/finx
make run      # build + launch

WebAssembly

make web      # build → docs/index.{html,js,wasm}
make serve    # build + serve at http://localhost:8000

Testing

make test             # unit tests (requires GoogleTest + libcurl)
make test-yfinance    # Yahoo Finance integration tests (requires pybind11 + yfinance)

Keyboard shortcuts

Shortcut	Action
Ctrl+N	Add new data stream
Ctrl+P	New plot window
Ctrl+S	Save config
Ctrl+E	Export PNG

On macOS Ctrl maps to Cmd.

Data sources

Source	Notes
CSV	Upload any delimiter-separated file; columns auto-typed
HTTP GET	JSON (dot-path extraction) or CSV response; configurable field map
Yahoo Finance	Ticker + period + interval; requires pybind11 integration
Formula	Arithmetic expression over bound stream fields (e.g. (a / b) * 100)

Formula engine

Formula streams let you derive new data from any loaded stream. Open the Formula Builder from the Data Streams panel, bind field aliases to columns in a source stream, then write an expression using those aliases.

Operators

Operator	Description
+ - * /	Add, subtract, multiply, divide
%	Modulo (remainder)
( )	Parentheses for grouping

Scalar functions

Evaluate independently per row. Arguments can be any expression.

Function	Description
abs(x)	Absolute value
sqrt(x)	Square root
pow(x, n)	x raised to the power n
log(x)	Natural logarithm (ln)
exp(x)	e raised to the power x
sin(x)	Sine (radians)
cos(x)	Cosine (radians)
tan(x)	Tangent (radians)

Window functions

Operate over a rolling window of N rows. Syntax: fn(alias, N) — alias must be a bound field name and N must be a plain integer. Window functions produce NaN for the first N−1 warmup rows; these appear as natural gaps in the plot rather than spurious zeros.

Function	Description	Notes
sma(alias, N)	Simple moving average over N periods	NaN for first N−1 rows
ema(alias, N)	Exponential moving average	Seeded with SMA; k = 2/(N+1)
stddev(alias, N)	Rolling population standard deviation	NaN for first N−1 rows
rmin(alias, N)	Rolling minimum over N periods	NaN for first N−1 rows
rmax(alias, N)	Rolling maximum over N periods	NaN for first N−1 rows
roc(alias, N)	Rate of change vs N periods ago	(v[i]−v[i−N]) / v[i−N] × 100

Example expressions

Expression	What it computes
(a / b) * 100	a as a percentage of b
close - sma(close, 20)	Distance from 20-day moving average
(close - sma(close, 20)) / stddev(close, 20)	Z-score vs 20-day window
ema(close, 12) - ema(close, 26)	MACD line
close / sma(close, 200) - 1	% above/below 200-day moving average
roc(close, 1)	Daily return %
rmax(high, 52) - rmin(low, 52)	52-period true range