diff --git a/Experimental Mouse Data Analysis.ipynb b/Experimental Mouse Data Analysis.ipynb index b008341..4be1df4 100644 --- a/Experimental Mouse Data Analysis.ipynb +++ b/Experimental Mouse Data Analysis.ipynb @@ -1118,13 +1118,13 @@ }, { "cell_type": "code", - "execution_count": 238, + "execution_count": 279, "metadata": {}, "outputs": [], "source": [ "from sklearn.cluster import KMeans\n", "from sklearn.preprocessing import StandardScaler\n", - "from sklearn import *" + "from sklearn.metrics import *" ] }, { @@ -1171,7 +1171,7 @@ }, { "cell_type": "code", - "execution_count": 241, + "execution_count": 308, "metadata": {}, "outputs": [], "source": [ @@ -1180,24 +1180,24 @@ }, { "cell_type": "code", - "execution_count": 242, + "execution_count": 309, "metadata": {}, "outputs": [ { "output_type": "execute_result", "data": { "text/plain": [ - "" + "" ] }, "metadata": {}, - "execution_count": 242 + "execution_count": 309 }, { "output_type": "display_data", "data": { "text/plain": "
", - "image/svg+xml": "\r\n\r\n\r\n\r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n\r\n", + "image/svg+xml": "\r\n\r\n\r\n\r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n\r\n", "image/png": "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\n" }, "metadata": { @@ -1213,7 +1213,7 @@ }, { "cell_type": "code", - "execution_count": 243, + "execution_count": 310, "metadata": {}, "outputs": [ { @@ -1233,24 +1233,24 @@ }, { "cell_type": "code", - "execution_count": 247, + "execution_count": 311, "metadata": {}, "outputs": [ { "output_type": "execute_result", "data": { "text/plain": [ - "" + "" ] }, "metadata": {}, - "execution_count": 247 + "execution_count": 311 }, { "output_type": "display_data", "data": { "text/plain": "
", - "image/svg+xml": "\r\n\r\n\r\n\r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n\r\n", + "image/svg+xml": "\r\n\r\n\r\n\r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n \r\n\r\n", "image/png": "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\n" }, "metadata": { @@ -1266,14 +1266,14 @@ }, { "cell_type": "code", - "execution_count": 245, + "execution_count": 312, "metadata": {}, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ - "0.3164176254356233\n" + "0.3164176254356233\n0.5427350983278696\n" ] } ], @@ -1281,12 +1281,14 @@ "ari_kmeans = adjusted_rand_score(colour_list, kmeans.labels_)\n", "print(ari_kmeans)\n", "#acc_score = accuracy_score(ground_truth, kmeans.labels_)\n", - "#print(acc_score)" + "#print(acc_score)\n", + "c_score = metrics.completeness_score(colour_list, kmeans.labels_)\n", + "print(c_score)" ] }, { "cell_type": "code", - "execution_count": 278, + "execution_count": 313, "metadata": {}, "outputs": [ { @@ -1294,55 +1296,105 @@ "name": "stdout", "text": [ "0.3164176254356233\n", + "0.5427350983278696\n", "0.2979396262577863\n", + "0.5299947088746245\n", "0.3164176254356233\n", + "0.5427350983278695\n", "0.2979396262577863\n", + "0.5299947088746244\n", "0.3164176254356233\n", + "0.5427350983278695\n", "0.2183488987664186\n", + "0.4731471416677808\n", "0.2522602375465343\n", + "0.5216642869861097\n", "0.2979396262577863\n", + "0.5299947088746245\n", "0.3164176254356233\n", + "0.5427350983278695\n", "0.3164176254356233\n", + "0.5427350983278695\n", "0.24191229331416247\n", + "0.4793723140812728\n", "0.3164176254356233\n", + "0.5427350983278695\n", "0.3164176254356233\n", + "0.5427350983278696\n", "0.317518583096265\n", + "0.5842337767212318\n", "0.2522602375465343\n", + "0.5216642869861099\n", "0.25466574503801714\n", + "0.4909772064840389\n", "0.2981540420114577\n", + "0.5763457410914344\n", "0.2979396262577863\n", + "0.5299947088746244\n", "0.2183488987664186\n", + "0.4731471416677808\n", "0.3164176254356233\n", + "0.5427350983278696\n", "0.3164176254356233\n", + "0.5427350983278695\n", "0.3164176254356233\n", + "0.5427350983278696\n", "0.2522602375465343\n", + "0.5216642869861097\n", "0.24191229331416247\n", + "0.4793723140812727\n", "0.24191229331416247\n", + "0.4793723140812728\n", "0.3010007322431047\n", + "0.5120528625709914\n", "0.2979396262577863\n", + "0.5299947088746245\n", "0.2522602375465343\n", + "0.5216642869861097\n", "0.3164176254356233\n", + "0.5427350983278695\n", "0.3164176254356233\n", + "0.5427350983278696\n", "0.2979396262577863\n", + "0.5299947088746245\n", "0.3164176254356233\n", + "0.5427350983278696\n", "0.3164176254356233\n", + "0.5427350983278696\n", "0.25466574503801714\n", + "0.5098941084715454\n", "0.3164176254356233\n", + "0.5427350983278695\n", "0.24315652261104986\n", + "0.5044336508279111\n", "0.3164176254356233\n", + "0.5427350983278696\n", "0.3164176254356233\n", + "0.5427350983278696\n", "0.2721311475409836\n", + "0.5430666005771546\n", "0.25466574503801714\n", + "0.4909772064840388\n", "0.2522602375465343\n", + "0.5216642869861099\n", "0.2979396262577863\n", + "0.5299947088746243\n", "0.3164176254356233\n", + "0.5427350983278696\n", "0.2979396262577863\n", + "0.5299947088746244\n", "0.3164176254356233\n", + "0.5427350983278695\n", "0.3164176254356233\n", + "0.5427350983278696\n", "0.2522602375465343\n", + "0.5216642869861097\n", "0.2979396262577863\n", + "0.5299947088746244\n", + "0.3164176254356233\n", + "0.5427350983278695\n", "0.3164176254356233\n", - "0.3164176254356233\n" + "0.5427350983278695\n" ] } ], @@ -1352,15 +1404,47 @@ " kmeans = KMeans(init=\"random\",n_clusters=6,n_init=10, max_iter=300, random_state=i*12+1)\n", " kmeans.fit(features)\n", " ari_kmeans = adjusted_rand_score(colour_list, kmeans.labels_)\n", - " print(ari_kmeans)" + " print(ari_kmeans)\n", + " c_score = metrics.completeness_score(colour_list, kmeans.labels_)\n", + " print(c_score)" ] }, { "cell_type": "code", - "execution_count": null, + "execution_count": 320, "metadata": {}, - "outputs": [], - "source": [] + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + "0.186480707936279\n0.32129820368474604\n0.16574030596851375\n0.24430063365576732\n0.22599736236274487\n0.28985598502257726\n0.23169710363478055\n0.2965313718255232\n0.21281215608478468\n0.23282149174131248\n0.31374114936393316\n0.1777214647695583\n0.24066938419061198\n0.17003674668492721\n0.2342058236244116\n0.375013898773401\n0.3349236381454569\n0.287136086274843\n0.33876792499972336\n0.28334616554756303\n0.21375625978513302\n0.21448732609637042\n0.1740265679695563\n0.23384339618103658\n0.26129923243607434\n0.1732955016583188\n0.22599736236274495\n0.21815132854445318\n0.2999045361451192\n0.192000164802131\n0.2560706136987088\n0.19563327289829416\n0.2641931670028276\n0.24684024799731796\n0.20844798167738407\n0.2737224780508776\n0.2611189480298907\n0.24934896798694917\n0.1320011935534592\n0.29639734632815334\n0.34101670121278715\n0.16618053415126452\n0.2990709861670738\n0.26177221360175246\n0.18339031168967446\n0.2561187312385306\n0.19058493856711242\n0.2048636385996851\n0.17246195168027342\n0.15422841107113155\nMean of Completeness Score = 0.23966643830951084\n" + ] + } + ], + "source": [ + "from random import randint\n", + "apple_pie = []\n", + "for i in np.arange(50):\n", + " # Nah this isn't good because it's still clustering based on closeness. \n", + " #kmeans = KMeans(init=\"random\",n_clusters=6,n_init=1, max_iter=1, random_state=i*12+1)\n", + " #kmeans.fit(features)\n", + "\n", + " # Use rand int to generate a list\n", + " rand_array = []\n", + " for j in np.arange(37):\n", + " rand_array.append(randint(0,5))\n", + "\n", + " # Confirms ari = 0 for random\n", + " ari_kmeans = adjusted_rand_score(rand_array, kmeans.labels_)\n", + " print(ari_kmeans)\n", + " #\n", + " c_score = metrics.completeness_score(rand_array, kmeans.labels_)\n", + " apple_pie.append(c_score)\n", + " print(c_score)\n", + "\n", + "print('Mean of Completeness Score = ', np.mean(apple_pie))" + ] } ], "metadata": {