Skip to content

Instantly share code, notes, and snippets.

@bdach

bdach/Choosing cutoff skin element sizes.ipynb

Last active October 5, 2023 20:25
Download ZIP

Revisions

  1. bdach revised this gist Oct 5, 2023. 1 changed file with 27 additions and 11 deletions.
    38 changes: 27 additions & 11 deletions Choosing cutoff skin element sizes.ipynb
    View file
    Original file line number Diff line number Diff line change
    @@ -1587,6 +1587,14 @@
    "plot_cumulative_distribution(ecdf)"
    ]
    },
    {
    "cell_type": "markdown",
    "id": "73115fa7",
    "metadata": {},
    "source": [
    "As it turns out, [some skins opt to put `sliderfollowcircle` into `sliderb` itself](https://discord.com/channels/188630481301012481/188630652340404224/1159576312244088902), so we should probably match sizing for both. This is a bit of a time paradox situation, since we haven't _done_ `sliderfollowcircle` yet in this notebook and I'm editing this in post after doing everything and receiving feedback, but anyways - I know I'm gonna choose 384x384 for that one, so let's do the same here."
    ]
    },
    {
    "cell_type": "code",
    "execution_count": 51,
    @@ -1596,7 +1604,7 @@
    {
    "data": {
    "text/plain": [
    "0.03754439370877727"
    "0.0053272450532724225"
    ]
    },
    "execution_count": 51,
    @@ -1605,7 +1613,7 @@
    }
    ],
    "source": [
    "1 - ecdf[256, 256]"
    "1 - ecdf[384, 384]"
    ]
    },
    {
    @@ -1680,6 +1688,14 @@
    "plot_cumulative_distribution(ecdf)"
    ]
    },
    {
    "cell_type": "markdown",
    "id": "384afb7d",
    "metadata": {},
    "source": [
    "Let's keep all of the `sliderb` piece limits the same, just for consistency."
    ]
    },
    {
    "cell_type": "code",
    "execution_count": 55,
    @@ -1698,7 +1714,7 @@
    }
    ],
    "source": [
    "1 - ecdf[256, 256]"
    "1 - ecdf[384, 384]"
    ]
    },
    {
    @@ -1864,7 +1880,7 @@
    "id": "aac779ea",
    "metadata": {},
    "source": [
    "### default-N"
    "### `default-N`"
    ]
    },
    {
    @@ -1957,7 +1973,7 @@
    "id": "5da7af89",
    "metadata": {},
    "source": [
    "Just for consistency..."
    "Generally I was going to do 256x256 for consistency with the hitcircle elements. However, [hitcircle numbers are scaled down by a constant of 0.8x](https://github.com/ppy/osu/blob/4d315cb718e6d747340d82a84400a5caac973af7/osu.Game.Rulesets.Osu/Skinning/Legacy/OsuLegacySkinTransformer.cs#L142-L146), so we'll do 256 / 0.8 = 320x320:"
    ]
    },
    {
    @@ -1969,7 +1985,7 @@
    {
    "data": {
    "text/plain": [
    "0.0012351397251814111"
    "0.00030878493129538054"
    ]
    },
    "execution_count": 66,
    @@ -1978,7 +1994,7 @@
    }
    ],
    "source": [
    "1 - ecdf[256, 256]"
    "1 - ecdf[320, 320]"
    ]
    },
    {
    @@ -2819,11 +2835,11 @@
    "| `sliderstartcircleoverlay` | 128x128 | 15.51% | 256x256 | 0.26% |\n",
    "| `sliderendcircleoverlay` | 128x128 | 4.90% | 256x256 | 0.72% |\n",
    "| `reversearrow` | 128x128 | 36.47% | 256x256 | 0.19% |\n",
    "| `sliderb` | 128x128 | 40.89% | 256x256 | 3.75% |\n",
    "| `sliderb-nd` | 128x128 | 3.57% | 256x256 | 3.57% |\n",
    "| `sliderb-spec` | 128x128 | 0.00% | 256x256 | 0.00% |\n",
    "| `sliderb` | 128x128 | 40.89% | 384x384 | 0.53% |\n",
    "| `sliderb-nd` | 128x128 | 3.57% | 384x384 | 3.57% |\n",
    "| `sliderb-spec` | 128x128 | 0.00% | 384x384 | 0.00% |\n",
    "| `sliderfollowcircle` | 308x308 | 4.04% | 384x384 | 1.52% |\n",
    "| `default-N` | 128x128 | 13.33% | 256x256 | 0.12% |\n",
    "| `default-N` | 128x128 | 13.33% | 320x320 | 0.03% |\n",
    "| `taikohitcircle` | 128x128 | 8.61% | 160x160 | 3.84% |\n",
    "| `taikobigcircle` | 128x128 | 9.22% | 160x160 | 2.99% |\n",
    "| `taikohitcircleoverlay` | 128x128 | 9.60% | 160x160 | 2.71% |\n",
  2. bdach created this gist Oct 5, 2023.
    2,859 changes: 2,859 additions & 0 deletions Choosing cutoff skin element sizes.ipynb
    View file
    2,859 additions, 0 deletions not shown because the diff is too large. Please use a local Git client to view these changes.