A scatter plot is a two-axis chart that places each x,y pair as one point, with the x value controlling horizontal position and the y value controlling vertical position. The Scatter Plot Maker turns pasted numeric rows of that same x,y shape into a downloadable SVG chart entirely in your browser, handling up to 200 non-empty rows with explicit valid, invalid, and over-limit counts so you always know what made it onto the canvas and what did not. It uses independent linear scales for each axis, so larger numeric gaps occupy more space and unequal spacing stays unequal, which is the same relationship a TI-84 STAT PLOT shows but produced from a copy-paste instead of poking through menus. The output is a standalone 800 by 500 SVG file with a title, grid, axis labels, and labeled point markers, and the on-screen preview and download come from the same percent-encoded SVG string so what you see is what you save.
Browser vs TI-84 for Scatter Plots
The TI-84 builds a scatter plot through its STAT editor: you enter lists L1 and L2, open STAT PLOT, choose the scatter icon, set the window, and press GRAPH. It is reliable, but it ties up the calculator, the screen is small, and exporting the chart for a report usually means a USB cable or a phone photo. A browser tool takes the same L1, L2 pairs you would otherwise enter on the calculator and reads them as comma-separated rows, so you never retype the numbers. The trade-off is that the browser tool is visualization-only: it does not compute correlation, fit a regression line, or run a hypothesis test. If your statistics class needs those values, you still run them on the calculator or a stats package. For a clean labeled chart that you can drop into a slide or a lab report, the browser workflow is shorter.
| Task | TI-84 STAT PLOT | Scatter Plot Maker |
|---|---|---|
| Where numbers are entered | Calculator lists L1, L2 | Pasted text in the page |
| Chart output | On-screen graph only | 800 by 500 SVG download |
| Axis window | Manual ZOOM or WINDOW | Auto from min and max of valid points |
| Trend line | Optional STAT then LinReg | Not included |
| Sharing the result | Cable transfer or phone photo | Direct file download |
| Statistical summary | r, regression, residuals, tests | None |
Preparing Your x,y Pairs Correctly
Every non-empty row must contain exactly two comma-separated numbers, and the rest is up to you. Signed values, decimals, plain zero, and scientific notation such as 1.5e-3 are all accepted. Blank lines are skipped without being counted as errors. Anything else gets flagged as invalid and skipped without becoming a zero point or moving to an axis. Negative zero is normalized to zero so -0 and 0 are treated the same. The parser is intentionally strict so the chart reflects only what you intended, and the summary will tell you exactly which rows did not make it on.
| Row example | Outcome |
|---|---|
| 1.5, 2.5 | Valid point |
| 1.5,2.5 | Valid point (whitespace ignored) |
| 1.5e3, -2 | Valid point |
| 0, 0 | Valid point |
| -0, 4 | Valid point (normalized to 0, 4) |
| blank line | Skipped silently |
| 5 kg, 3 | Invalid (unit text) |
| 1, 2, 3 | Invalid (extra comma) |
| 1e15, 2 | Invalid (above 1e12 magnitude) |
| NaN, 3 | Invalid (non-finite token) |
If your numbers live in Excel, you can convert two columns into comma-separated rows by exporting as CSV and keeping just the two columns of interest. For a fuller prep walkthrough on three-variable data, the Excel scatter plot with three variables guide shows the column selection steps in detail.
Building the Scatter Plot From Your Pasted Data
- Open the Scatter Plot Maker and type a title into the title field, or leave it blank for an untitled chart.
- Paste one x,y pair per non-empty line into the data field. Signed values, decimals, plain zero, and scientific notation are all fine.
- Click generate. The tool parses up to the first 200 non-empty rows, leaving everything beyond that counted as over-limit rather than mixed into the invalid count.
- Read the three summary numbers: valid points drawn, invalid rows skipped, and over-limit rows skipped. If valid is zero, the chart is cleared and an error appears.
- Check both axis labels to confirm the displayed range matches what you expect; zero is not auto-included.
- Hover or inspect representative points in the preview to confirm the labeled x,y values match your input.
- Download the standalone SVG. The download and the preview come from the same percent-encoded SVG string, so editing the title or data disables the previous download.
Reading the Row Counts and Axis Labels
After generation, the summary panel shows three separate counts: valid points, invalid rows, and over-limit rows. The valid count is the number of points actually drawn on the chart. The invalid count is the number of non-empty rows that the parser could not read as two finite numbers within the magnitude limit. The over-limit count is the number of non-empty rows that the parser never even attempted, because they appeared after the first 200. These counts always display, even when zero, so you can spot a paste that mostly worked and a paste that silently dropped data.
The axis labels expose the displayed range: the smallest x value is anchored to the left, the largest to the right, and values between are placed in direct proportion to their numeric distance. Same rule on y, smallest at the bottom and largest at the top. Five reference lines per axis make the numeric boundaries easy to read. The axes do not automatically include zero, so before you interpret the spread of a cluster, look at the labels and confirm whether zero is actually on the chart. A cluster that looks tight at the top of the chart may sit far from zero if the y window starts at 90.
Edge Cases: Identical Values, Scientific Notation, and Beyond 200 Rows
When every x value is the same number, the minimum and maximum would be equal and the normalized fraction would divide by zero. The tool handles that by expanding the x domain symmetrically around the constant, using ten percent of the absolute constant or one unit, whichever is larger. The same rule applies independently to a constant y value. A single valid point therefore sits in the center of both axes instead of failing the chart, which matters for calibration samples where one variable is held fixed.
Scientific notation is parsed by the same rules as plain numbers, so 1.5e3, 2.5e-2, and 0.0000025 are all valid inputs. Numbers with absolute magnitude above 1e12 are rejected as invalid, since coordinate precision in an 800 by 500 SVG becomes meaningless past that range. The total paste is capped at 50,000 UTF-16 code units, which keeps the page responsive. If your dataset exceeds 200 rows, only the first 200 are checked, and every subsequent non-empty row is reported as over-limit. They are not folded into the invalid count, because they were deliberately not parsed. Nothing is silently truncated or implied to be all used.
When the TI-84 Still Makes Sense
The browser tool deliberately stops at visualization. It does not draw a trend line, calculate correlation r, fit a regression equation, or run a t-test. If your task is to produce a fitted line and the r² value for a lab report, the TI-84 STAT then CALC menu, or any dedicated statistics package, is the right place to do that, and the chart can be exported afterward if you need the picture.
The tool also has limits that do not match a statistics class. Categorical color, logarithmic axes, accessible point labels, and dense overlays are not supported, so if your dataset has thousands of points or needs custom styling, switch to a plotting library such as matplotlib or plotly and retain the original numeric rows as your source of truth. Keep the original rows for analysis, uncertainty, reproducibility, and exact values, since the chart rounds coordinates to three decimal places for display.
All parsing, coordinate calculation, SVG assembly, preview rendering, and download preparation happen locally in the current browser tab. No points and no title are uploaded, which makes the tool safe to use with classroom data or measurements that should not leave your machine. Before you call a chart finished, verify both axis ranges, the three row counts, and a few point titles. Overlapping points can hide one another, especially when identical pairs repeat. If you want a denser view of the same x,y rows, the scatter plot to heatmap converter turns the same data into a density grid for visual inspection.
The underlying file format follows the W3C SVG 2 specification, so the chart is a valid standalone vector file you can open in any modern browser or vector editor and resize without losing sharpness.