Skip to content

Instantly share code, notes, and snippets.

@NaimKabir
Last active November 14, 2020 13:58
Show Gist options
  • Save NaimKabir/ba7516fb624109ae7e01c85c0c12584e to your computer and use it in GitHub Desktop.
Save NaimKabir/ba7516fb624109ae7e01c85c0c12584e to your computer and use it in GitHub Desktop.
Display the source blob
Display the rendered blob
Raw
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Simple process: let's say you have two populations. Split-ticket voters, and voters who vote using the straight ticket.\n",
"\n",
"We assume Republican straight-ticket voters always vote for Trump.\n",
"\n",
"Split-ticket voters instead have some chance of voting for Trump $p({trump})$, where $p({trump})$ is correlated with the straight-ticket percentage of Trump voters."
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [],
"source": [
"correlation = 0.9\n",
"p_trump = lambda percent_r: correlation*percent_r"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Great now let's simulate a bunch of precincts with different fractions of R voters.\n",
"\n",
"We'll assume that all R-voters voted Trump."
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [],
"source": [
"import numpy as np\n",
"num_precincts = 1000 # EDIT: let's deal with 1000 precincts, to reduce noisiness in the graph\n",
"r_percentages = np.random.rand(num_precincts) # modeling the fraction of R voters in a precint, randomly from [0-1]"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Now let's simulate some per-precinct votes and collect % of Trump votes from non-Republican candidates.\n",
"\n",
"This can be sampled a a bunch of weighted coinflips with the probability $p({trump})$--i.e, sampling from a binomial distribution.\n",
"\n",
"In this correlated case, the $p({trump})$ will be dependent on the straight-ticket % of Republicans in the precinct, though. "
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {},
"outputs": [],
"source": [
"# let's say we're dealing with a fixed size of non-republican voters in each precinct\n",
"non_r_pop_size = 500\n",
"split_trump_percentages = np.zeros(num_precincts)\n",
"# simulating\n",
"for i, percent in enumerate(r_percentages):\n",
" split_votes = np.random.binomial(non_r_pop_size, p_trump(percent)) \n",
" split_trump_percentages[i] = split_votes / non_r_pop_size\n"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"import matplotlib.pyplot as plt\n",
"import seaborn as sns\n",
"sns.set()\n",
"\n",
"diffs = split_trump_percentages - r_percentages\n",
"plt.scatter(r_percentages*100, diffs*100) \n",
"plt.xlabel('% republican straight-ticket votes')\n",
"plt.ylabel('% individual trump votes - % straight ticket votes')\n",
"plt.savefig('corr-diff')"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3",
"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": "3.7.7"
}
},
"nbformat": 4,
"nbformat_minor": 4
}
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment