Scatter Plot Maker surfaces three explicit row counts — valid points, invalid rows, and over-limit rows — along with five labeled reference lines per axis and a per-circle text title showing the formatted x and y values, so the entire post-generation verification happens inside the generated SVG and the on-page preview without opening a second tool. After you click Generate, the first thing to check is the summary block directly above the chart: it tells you exactly how many of your pasted x,y pairs were plotted, how many were rejected by the strict parser, and how many were skipped because they sat beyond the first 200 non-empty lines. Next, read the five reference lines along each axis — those labels are the actual numeric range the chart used, because the axes do not automatically include zero. Then open the SVG (or hover a circle in some SVG viewers) and read the title text inside a few of the blue circles to confirm individual coordinates against your source data. The downloaded SVG is generated from the same percent-encoded string as the on-page preview, so once those three checks pass, the file you save is the file you see.
What the Three Row Counts Reveal
The summary block above the chart separates three buckets that other tools often conflate. Understanding which bucket a row falls into is the foundation of verification: every non-empty row lands in exactly one of them, and the totals add up to the number of non-empty rows you pasted. If the valid count is zero, the chart area clears and the tool reports an error rather than rendering a misleading empty plot. If the over-limit count is non-zero, the rows past row 200 were deliberately not parsed; they are not silently truncated and mixed into the valid total, so the valid number is not a false promise of "all your data."
| Count | What it represents | What to do when it is non-zero |
|---|---|---|
| Valid points | Non-empty rows that became blue circles in the preview | Compare to your intended dataset size up to the 200-row boundary |
| Invalid rows | Non-empty rows the strict parser rejected | Inspect each rejected row for stray commas, unit text, hex values, NaN, or magnitudes above 1e12 |
| Over-limit rows | Non-empty rows past the first 200 the parser did not check | Decide whether to work with the first 200 or split the dataset into smaller pieces |
Invalid rows never become zero points or jump to an axis — they are simply skipped and counted. Over-limit rows are kept separate from invalid rows specifically because the parser never looked at them; conflating the two would hide which category a stray row actually belonged to.
How to Check a Scatter Plot Result in Scatter Plot Maker
A focused verification routine takes only a minute once you know what to look at. Open the Scatter Plot Maker page and walk through these concrete steps after every Generate click.
- Read the valid-points count and compare it to the number of pairs you intended to plot. If it disagrees, check whether your paste included header lines, unexpected blank rows, or trailing commas.
- Read the invalid-rows count. For each reported invalid row, return to your source and look for stray text, unit suffixes, extra commas, hexadecimal values, NaN, or magnitudes above 1e12.
- Read the over-limit count. If you pasted more than 200 non-empty rows, the rows past 200 were deliberately skipped, so decide whether to trust the first 200 or split the dataset.
- Scan the five axis reference lines on the horizontal axis. The leftmost label is your minimum x, the rightmost is your maximum x, and the three middle labels let you estimate where a cluster sits on the numeric scale.
- Repeat the scan on the vertical axis. The bottom label is your minimum y and the top label is your maximum y.
- Open the downloaded SVG (or the preview in an SVG-aware viewer) and read the text title inside three to five circles. Each title contains the exact formatted x and y values used for that point.
- Cross-check those titles against two or three rows in your source data. If a circle's title disagrees with your row, the parser applied a normalization — negative zero becomes zero — or your row was miscounted in the summary.
- Confirm that the displayed axis range matches what you expected. If your data includes 0 and the minimum label is not 0, remember the axes do not auto-include zero and the chart range is driven solely by min and max.
Reading the Axis Range Labels on the Preview
Scatter Plot Maker uses independent linear scales. The smallest x value in your data maps to the left plot boundary and the largest x maps to the right, with all intermediate values placed in direct proportion to their numeric distance. The same rule applies to y on the vertical axis, with the smallest y at the bottom and the largest at the top. Because the labels are printed on the chart itself, verification does not require a second calculation: read the left and right labels for x, read the bottom and top labels for y, and you have the displayed domain.
| Axis reading | What the label shows | Why it matters |
|---|---|---|
| Minimum label | Smallest value that maps to the chart edge | Reveals whether your data has been shifted away from zero |
| Middle labels | Three evenly placed intermediate values | Lets you estimate where a cluster sits on the numeric scale |
| Maximum label | Largest value that maps to the opposite edge | Confirms the spread the chart actually rendered |
| Center reference line | Midpoint of the displayed range | Use it to spot-check a few plotted points |
| Constant-axis expansion | Symmetric domain centered on the single value | Tells you the tool handled a constant axis instead of failing |
Two cases deserve special attention during verification. If every x value is identical, the tool expands the x domain symmetrically around that constant — ten percent of the absolute constant or one unit, whichever is larger — and centers every point horizontally; the same rule applies independently to y. The visible label on that axis shows a single repeated value, which is the visual cue that the constant-axis rule fired rather than a divide-by-zero error. Negative zero in your data is normalized to zero, so a row of -0,0 produces a circle at exactly (0,0) and a non-zero minimum label will still read 0.
Inspecting Individual Points Through SVG Titles
Every circle in the generated SVG contains a title element holding the exact formatted x and y values used to position that point. In an SVG-aware viewer such as a modern browser, you can read the title by opening the file directly or by hovering when the viewer supports tooltips. The MDN SVG tutorial documents this title element, and it is part of the standard SVG 2 specification, so any tool that opens the file will render it. This makes point-level verification possible: pick three rows from your source, find the corresponding circles in the chart, and read the title text. If your row was 3.5,7.25 the title shows those two numbers formatted by the tool with up to three decimal places of coordinate rounding, which is precise enough to confirm a point landed in the right cluster but not precise enough to recover the original floating-point value — keep your source rows for any downstream analysis.
Confirming the Download Matches the Preview
The on-page preview and the downloaded SVG are produced from the same percent-encoded SVG data URI. There is no second render path that could diverge from the first, so the chart you see before clicking download is byte-for-byte the chart you save after clicking download — including the title, the grid, the axis labels, the reference lines, and every circle with its embedded title text. That property lets verification be done on the preview: if the preview looks right, the file will look right. The inverse also holds: editing the title field or the data block clears the generated string and disables the old download, so re-verify after any change before reusing a previously downloaded file. Coordinates are rounded to three decimal places in the SVG, which means the file is a visualization, not a precision-preserving data format; store your original rows separately.
When Verification Suggests You Need a Different Tool
Verification can also tell you when Scatter Plot Maker is the wrong tool for the job. The chart does not calculate correlation, regression, a trend line, confidence intervals, error bars, density, clusters, or statistical significance — it only plots supplied pairs on linear axes. If your verification checklist demands an R² value, a fitted slope, or a hypothesis test, generate the SVG for visual inspection but run the analysis in a dedicated statistical package against your source rows. For dense data with overlapping points, categorical color, accessible labeling, logarithmic scales, or publication styling, the same applies: keep the SVG for quick visual checks and reach for a plotting library when the deliverable demands more. For a closer look at the kind of mistakes these verification steps catch, the guide on avoiding common pitfalls with a scatter plot maker walks through the same checks from the opposite direction and is a useful companion read.