ASPP/2025-plovdiv-scientific-patterns
0
0
Fork 17
Code Issues 6 Pull requests 9 Projects Releases Packages Wiki Activity Actions

adds step 4 exercise

Browse source
This commit is contained in:
Lisa Schwetlick 2025-09-24 10:15:00 +02:00
parent 5455d9be1d
commit e96a56f177
1 changed files with 18 additions and 18 deletions
Download patch file Download diff file Expand all files Collapse all files

36
notebooks/walker/Step_4_break_out_the_next_step_probability/Step_4_break_out_the_next_step_probability.ipynb
View file

@ -10,7 +10,7 @@
}
},
"source": [
"# Break out the next step proposal\n",
"# How can we break out the next step proposal?\n",
"\n",
"- We want a separate module for `next_step_proposals`, living in a separate file\n",
" - a Gaussian one (it's the one currently in the Walker code), called `gaussian_next_step_proposal`\n",
@ -102,27 +102,11 @@
"plot_trajectory(trajectory, walker.context_map)"
]
},
{
"cell_type": "markdown",
"metadata": {
"nteract": {
"transient": {
"deleting": false
}
}
},
"source": [
"# 2. Implement the changes in the code\n",
"1. In a new module `next_step_proposals`, write two functions for two different next step proposals: a Gaussian one (it's the one currently in the Walker code) and a square one (you can find an example below)\n",
"2. Modify the constructor of Walker to take a `next_step_proposal` function and a `next_step_proposal_arguments` dictionary as an input\n",
"3. See if the dream code you wrote works with your implementation! (Don't forget to restart the kernel)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# What the square next step proposal could look like\n",
"# What the square next step proposal could look like (info)\n",
"\n",
"This is an example of what the square next step proposal function could look like.\n",
"\n",
@ -175,6 +159,22 @@
"plt.colorbar()"
]
},
{
"cell_type": "markdown",
"metadata": {
"nteract": {
"transient": {
"deleting": false
}
}
},
"source": [
"# 2. Implement the changes in the code (optional)\n",
"1. In a new module `next_step_proposals`, write two functions for two different next step proposals: a Gaussian one (it's the one currently in the Walker code) and a square one (you can find an example below)\n",
"2. Modify the constructor of Walker to take a `next_step_proposal` function and a `next_step_proposal_arguments` dictionary as an input\n",
"3. See if the dream code you wrote works with your implementation! (Don't forget to restart the kernel)"
]
},
{
"cell_type": "code",
"execution_count": 7,