{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "# Counting cells with a Poisson noise model\n", "A core component of all calibration models are the distribution that describes the spread of measurements, also known as the _noise model_.\n", "If the noise models already provided by `calibr8` don't match for your application, you can implement your own noise model by following the example below." ] }, { "cell_type": "code", "execution_count": 1, "metadata": {}, "outputs": [], "source": [ "import calibr8\n", "import numpy\n", "import scipy.stats\n", "from matplotlib import pyplot" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Implementing a noise model\n", "Noise models must inherit from `calibr8.DistributionMixin` and implement at least the following:\n", "\n", "* A `scipy_dist` class attribute to identify a SciPy probability distribution.\n", "* A `to_scipy` method to map the tuple of parameters returned by the `predict_dependent` method to a dictionary of keyword arguments for `scipy_dist`.\n", "\n", "Note that the class inheriting the noise model is responsible to implement `predict_dependent` such that the mapping methods apply.\n", "We encourage everybody to [write unit tests for custom noise models](https://github.com/JuBiotech/calibr8/blob/v6.2.0/calibr8/tests.py#L145)." ] }, { "cell_type": "code", "execution_count": 2, "metadata": {}, "outputs": [], "source": [ "class PoissonNoise(calibr8.DistributionMixin):\n", " scipy_dist = scipy.stats.poisson\n", "\n", " @staticmethod\n", " def to_scipy(*params):\n", " return dict(mu=params[0])\n", "\n", "\n", "# Note: A `PoissonNoise` model with optional PyMC support is already included in calibr8." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Using the noise model in a custom calibration model\n", "The custom noise model mixin can be used just like any other noise model.\n", "\n", "To set up a calibration model for cell dry weight vs. number of cells counted in a counting chamber, we can wire up the Poisson's only parameter as proportional to the cell dry weight." ] }, { "cell_type": "code", "execution_count": 3, "metadata": {}, "outputs": [], "source": [ "class CellCountCalibration(calibr8.ContinuousUnivariateModel, PoissonNoise):\n", " def __init__(self):\n", " super().__init__(\n", " independent_key=\"cell dry weight / g/L\",\n", " dependent_key=\"cells counted / -\",\n", " theta_names=[\"slope\"],\n", " )\n", "\n", " def predict_dependent(self, x, *, theta=None):\n", " if theta is None:\n", " theta = self.theta_fitted\n", " slope = theta[0]\n", " # Always return a tuple 👇\n", " return (slope * x,)\n", "\n", "\n", "ccc = CellCountCalibration()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Fitting the model to toy data\n", "As with other examples, we need some data to play with.\n", "Let's say we have some calibration data points with known cell dry weight and a corresponding cell count." ] }, { "cell_type": "code", "execution_count": 4, "metadata": {}, "outputs": [], "source": [ "# Ten known cell dry weights [g/L].\n", "cdw = numpy.linspace(1, 20, 10)\n", "\n", "rng = numpy.random.RandomState(20211122)\n", "# Using 0.5 as the slope gives us a clearly distinguishable y-axis later.\n", "obs = rng.poisson(lam=cdw * 0.5)\n", "\n", "_, history = calibr8.fit_scipy(\n", " ccc,\n", " independent=cdw,\n", " dependent=obs,\n", " theta_guess=[1],\n", " theta_bounds=[(0.1, 10)],\n", ")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Visualizing the fitted model\n", "For visualizing this model, we need to think about how to deal with the discrete dependent variable.\n", "Let's try a heatmap to discretize the y-axis, but use a high resolution on the x-axis to make it look continuous." ] }, { "cell_type": "code", "execution_count": 5, "metadata": {}, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "fig, ax = pyplot.subplots(dpi=200)\n", "\n", "xdense = numpy.linspace(0.01, 20, 1000)\n", "ydense = numpy.arange(0, 10)\n", "p = ccc.scipy_dist.pmf(ydense[:, None], mu=ccc.predict_dependent(xdense)[0][None, :])\n", "\n", "sm = ax.imshow(\n", " p[::-1, :],\n", " extent=(0.01, 20, 0.5, 10.5),\n", " aspect=\"auto\",\n", " interpolation=\"none\",\n", " cmap=pyplot.cm.Greens,\n", " vmax=1,\n", ")\n", "cbar = pyplot.colorbar(\n", " sm,\n", " cax=fig.add_axes([0.92, 0.126, 0.03, 0.7545]),\n", ")\n", "cbar.set_label(\"probability of observation / -\", rotation=270, labelpad=15)\n", "ax.scatter(cdw, obs, marker=\"x\")\n", "ax.set(\n", " ylabel=ccc.dependent_key,\n", " xlabel=ccc.independent_key,\n", " ylim=(0.5, None),\n", " xlim=(0, None),\n", ")\n", "\n", "pyplot.show()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Applying the model for UQ\n", "Assuming that from one sample we have counted cells twice, can we now state our uncertainty about the cell dry weight in this sample?" ] }, { "cell_type": "code", "execution_count": 6, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "With a prior belief that we can't have less than 0 and more than 20 g/L cell dry weight, \n", "and after observing 4 and 5 cells, our posterior belief is that the \n", "cell dry weight of our sample is in the [6.88, 18.20] g/L interval with a 90 % chance.\n" ] }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "def show_posterior(pst, yobs):\n", " print(\n", " \"With a prior belief that we can't have less than 0 \"\n", " \"and more than 20 g/L cell dry weight, \\n\"\n", " f\"and after observing {', '.join(map(str, yobs[:-1]))} and {yobs[-1]} cells, \"\n", " \"our posterior belief is that the \\ncell dry weight of our sample \"\n", " f\"is in the [{pst.hdi_lower:.2f}, {pst.hdi_upper:.2f}] g/L interval \"\n", " f\"with a {pst.hdi_prob * 100:.0f} % chance.\"\n", " )\n", "\n", " fig, ax = pyplot.subplots()\n", "\n", " pst_full = ccc.infer_independent(y=yobs, lower=0, upper=20)\n", " ax.plot(pst_full.eti_x, pst_full.eti_pdf, color=\"gray\")\n", " ax.plot(\n", " pst.hdi_x,\n", " pst.hdi_pdf,\n", " lw=5,\n", " label=f\"{pst.hdi_prob * 100:.0f} % HDI\",\n", " color=\"orange\",\n", " )\n", " ax.set(\n", " ylim=(0, None),\n", " ylabel=\"posterior probability / -\",\n", " xlim=(0, 20),\n", " xlabel=\"cell dry weight / g/L\",\n", " )\n", " ax.legend()\n", "\n", " pyplot.show()\n", "\n", "\n", "yobs = [4, 5]\n", "pst = ccc.infer_independent(y=yobs, lower=0, upper=20, ci_prob=0.9)\n", "show_posterior(pst, yobs)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Oh, this is not very accurate!\n", "Can we get a more accurate estimate of the cell dry weight by cell-counting multiple times?" ] }, { "cell_type": "code", "execution_count": 7, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "With a prior belief that we can't have less than 0 and more than 20 g/L cell dry weight, \n", "and after observing 4, 5, 4, 6, 3 and 7 cells, our posterior belief is that the \n", "cell dry weight of our sample is in the [8.98, 16.51] g/L interval with a 90 % chance.\n" ] }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "yobs2 = yobs + [4, 6, 3, 7]\n", "pst2 = ccc.infer_independent(y=yobs2, lower=0, upper=20, ci_prob=0.9)\n", "show_posterior(pst2, yobs2)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "As expected: More observations give more accurate results.\n", "\n", "## Conclusion\n", "In this example we have seen how to implement a custom noise model, and build a calibration model with it.\n", "\n", "Furthermore, this example demonstrated that key components of `calibr8` such as `fit_scipy` or `infer_independent` are agnostic to wheather the noise model is a continuous or discrete distribution.\n", "\n", "This specific application showed that `calibr8` brings everything needed to do uncertainty quantification for cell-counting experiments." ] }, { "cell_type": "code", "execution_count": 8, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Last updated: Wed, 15 Apr 2026\n", "\n", "Python implementation: CPython\n", "Python version : 3.13.12\n", "IPython version : 9.11.0\n", "\n", "calibr8 : 7.2.0\n", "matplotlib: 3.10.8\n", "numpy : 2.4.2\n", "platform : 1.0.8\n", "scipy : 1.17.1\n", "\n", "Watermark: 2.6.0\n", "\n" ] } ], "source": [ "%load_ext watermark\n", "%watermark -n -u -v -iv -w" ] } ], "metadata": { "kernelspec": { "display_name": "Python 3 (ipykernel)", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.13.12" } }, "nbformat": 4, "nbformat_minor": 4 }