Number of data points limited to 10,000

The plots of any function on the platform are rendered by your browser. The more data points have to be plotted and the more memory is consumed to load the data and render the plot. This might slow down the notebook significantly and result in a bad user experience. Eventually, your browser might even run out of memory.

To prevent this and ensure a good notebook performance the number of data points which are plotted is limited to 10,000 by default. These 10,000 data points are a random subset of the original dataset. In most cases the random sample should have a distribution and statistics very close to the full dataset. The distribution might become distorted with regard to min and max values, especially if the distribution is sparse at its ends. You might also miss on outliers!

If a plot was generated on only 10,000 points instead of the full data set, a warning will be printed above the plot. This also informs you about the size of the original dataset.

You can override this behavior and nevertheless plot all points. At the bottom of any plotting function appears an additional option Plot all … data points? if the selected dataset is larger than 10,000 points. You can tick that option to include all data points in the plot. Use this option carefully as there can be a big impact on the notebook performance.