Simple Plotting Tool
The purpose of SPT is to allow researchers to view raw data in a
graphical format. SPT can produce high-quality graphical displays
of data as .pdf, .ps and .png images. The goal is to encourage
better data analysis by providing a tool that enables data to be
easily viewed in a variety of different formats. SPT makes it easy
to create standard lines plots, scatter plots and bar charts. Using
SPT, it is also simple to create box plots, mean-value lines with
95% confidence intervals, and histograms to display distributions of
data. Additionally, SPT has support for displaying trees (a
standard data structure for many computer science applications).
Some features of SPT are:
- Many different types of plots, including:
- Scatter plots
- Scatter plots with labels and error bars
- Lines plots
- Lines plots with error bars showing 95% confidence
- Bubble plots
- Cumulative density functions
- Heat maps
- Box plots with 95% confidence intervals
- Stacked Barcharts
- Fanned Barcharts
- Grouped nominal plots (bar charts and boxplots)
- Sunburst trees diagrams (library version only)
- Arbitrary single variable functions (library version only)
- A simple graphical user interface for viewing plots
- Intelligent tick-mark placement
- Data is never obstructed by an axis (uses proper padding)
- Performs some simple data analysis (confidence intervals, box
statistics, histogram binning, etc.)
- An Objective Caml library interface
- A simple lisp-inspired plotting language
- Generates: .ps, .pdf or .png images
SPT is available either as a
library for the Objective Caml
language or as a binary which reads a simple (lisp-inspired)
plotting language. Future releases are expected to have bindings
allowing library access via C or C++ programs.
Warning: some samples may appear lower quality since png images are
not a vector-graphics format; they are rasterized.
Mean lines with error bars at 95% confidence intervals.
More lines with error bars.
Lines and points.
Grouped box plots
with a 95% confidence interval on the mean.
A histogram of
1,000 values chosen uniformly from the range 0-10mil.
plot of a randomly colored tree with random branching between 2-3
and a depth of 10.
A stacked (left) and
layered (right) bar chart of fictitious data.
- spt-it-out manual (updated 2011-03-03)
The current version is a beta.