lint

analysis/team_gen_analysis.ipynb

@@ -8,10 +8,11 @@
    "source": [
     "import matplotlib.pyplot as plt\n",
     "import seaborn as sns\n",
-    "import numpy as np \n",
-    "import pandas as pd \n",
+    "import numpy as np\n",
+    "import pandas as pd\n",
+    "\n",
     "sns.set()\n",
-    "plt.style.use('ggplot')\n"
+    "plt.style.use(\"ggplot\")"
    ]
   },
   {
@@ -220,15 +221,16 @@
     }
    ],
    "source": [
-    "from glob import glob \n",
+    "from glob import glob\n",
+    "\n",
     "data_folder = glob(\"../outputs/*\")\n",
     "\n",
     "tdfs = []\n",
     "\n",
     "for folder in data_folder:\n",
-    "    files=glob(folder + \"/*.csv\")\n",
+    "    files = glob(folder + \"/*.csv\")\n",
     "    tdf = pd.concat([pd.read_csv(f) for f in files])\n",
-    "    tdf['generation'] = folder.split('/')[-1]\n",
+    "    tdf[\"generation\"] = folder.split(\"/\")[-1]\n",
     "    tdfs.append(tdf)\n",
     "\n",
     "data = pd.concat(tdfs)\n",
@@ -262,8 +264,8 @@
     }
    ],
    "source": [
-    "n_poke_per_gen = data.groupby('generation')['0'].nunique().sort_values(ascending=False)\n",
-    "fig, ax = plt.subplots(1, figsize=(5,5), dpi=100)\n",
+    "n_poke_per_gen = data.groupby(\"generation\")[\"0\"].nunique().sort_values(ascending=False)\n",
+    "fig, ax = plt.subplots(1, figsize=(5, 5), dpi=100)\n",
     "\n",
     "ax.plot(np.arange(len(n_poke_per_gen)), n_poke_per_gen)\n",
     "ax.set_ylabel(\"Number of Unique Pokemon\")\n",
@@ -518,10 +520,10 @@
     }
    ],
    "source": [
-    "tmp = data.groupby(['0','generation'])['0'].count().reset_index(name='count') \n",
+    "tmp = data.groupby([\"0\", \"generation\"])[\"0\"].count().reset_index(name=\"count\")\n",
     "for gen in data.generation.unique():\n",
-    "    fig, ax = plt.subplots(1, figsize=(10,10), dpi=100)\n",
-    "    sns.barplot(data=tmp[tmp['generation']==gen], y='0', x='count', ax=ax)\n",
+    "    fig, ax = plt.subplots(1, figsize=(10, 10), dpi=100)\n",
+    "    sns.barplot(data=tmp[tmp[\"generation\"] == gen], y=\"0\", x=\"count\", ax=ax)\n",
     "    ax.set_title(f\"Generation: {gen}\")\n",
     "    fig.savefig(f\"{gen}.png\")"
    ]

hypotheses_pseudocode.md

@@ -7,31 +7,36 @@
 - Team size: 1
 - Total Pokemon: 151
 - Battle Mode: ???
-- Starting number of teams: 151*151
-- Number of Battles per pairing: ??? 
+- Starting number of teams: 151\*151
+- Number of Battles per pairing: ???
 - Generations: `0`
 - Mutation rate: `0`
 - Breeding: `Off`
 - Loss Shuffle: `Off`
 
-### Hypotheses: 
+### Hypotheses:
 
-Is there a decay function for the number of battles? The win ratio should converge which will give us a good understanding of the number of battles required. 
+Is there a decay function for the number of battles? The win ratio should
+converge which will give us a good understanding of the number of battles
+required.
 
-How does this change based on the battle-mode? Is there a "good enough" mode that doesn't take too long?
+How does this change based on the battle-mode? Is there a "good enough" mode
+that doesn't take too long?
 
-Do we find all the S tier pokemon from the fan rankings?
-Do we find all the D tier pokemon from the fan rankings?
+Do we find all the S tier pokemon from the fan rankings? Do we find all the D
+tier pokemon from the fan rankings?
 
-Plotting in 2d using the sprites. The first dimension is the fitness, second dimension:
-- Base stat total = sum (hp, attack, defence, special attack, special defence, special speed)
+Plotting in 2d using the sprites. The first dimension is the fitness, second
+dimension:
+
+- Base stat total = sum (hp, attack, defence, special attack, special defence,
+  special speed)
 - HP
 - Attack
 - Anything else
 
-Nearest k-nearest neighbours clusters using fitness values, what tiers have we found?
-
-
+Nearest k-nearest neighbours clusters using fitness values, what tiers have we
+found?
 
 ## 2v2 Pokemon Battles
 
@@ -47,26 +52,29 @@ Nearest k-nearest neighbours clusters using fitness values, what tiers have we f
 - Breeding: `Off`
 - Loss Shuffle: `On`
 
-### Hypotheses: 
+### Hypotheses:
 
-Are there synergestic pairs that have greater wins compared with the top tier solo pokemon?
+Are there synergestic pairs that have greater wins compared with the top tier
+solo pokemon?
 
 Are there pairings of Pokemon types that stand out? (fire and water)
 
-What is the base rate of synergy? Can we see clusters of all the [Type1&Type2] pairs or are there some stand out power couples?
+What is the base rate of synergy? Can we see clusters of all the [Type1&Type2]
+pairs or are there some stand out power couples?
 
 2d clusterplot: first dimension is fitness, second dimension:
+
 - Categorical Pairs by type
 - Sum of base stats
 
+Can we find a diminishing returns number of battle repeats that converge on an
+answer?
 
-Can we find a diminishing returns number of battle repeats that converge on an answer?
-
-Can we find a diminishing returns number of generations that converge on an answer?
-
-What happens with a high versus low mutation rate? to how quickly it finds an answer and how variable the result is. 
-
+Can we find a diminishing returns number of generations that converge on an
+answer?
 
+What happens with a high versus low mutation rate? to how quickly it finds an
+answer and how variable the result is.
 
 ## 6v6 Pokemon Battles
 
@@ -82,20 +90,13 @@ What happens with a high versus low mutation rate? to how quickly it finds an an
 - Breeding: `On`
 - Loss Shuffle: `Off`
 
-### Hypotheses: 
+### Hypotheses:
 
 Which is the best Pokemon team?
 
-Are the best teams very related to the top Pokemon, or are there plenty of wildcards?
+Are the best teams very related to the top Pokemon, or are there plenty of
+wildcards?
 
 Are the synergestic pairs in the top teams?
 
 What are the spread of types in the top teams?
-
-
-
-
-
-
-
-