Skip to content

Instantly share code, notes, and snippets.

@Raahul-Singh

Raahul-Singh/Point Biserial.ipynb

Created June 28, 2020 20:49
Show Gist options
  • Save Raahul-Singh/2e99856951b8b648508896a9cca067bb to your computer and use it in GitHub Desktop.
Save Raahul-Singh/2e99856951b8b648508896a9cca067bb to your computer and use it in GitHub Desktop.
Download ZIP
Display the source blob
Display the rendered blob
Raw
Point Biserial.ipynb
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Correaltion between Flare Production and AR Complexity"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"The purpose of this note book is to find the correlation, if any exists, between Flare Production and AR Complexity."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"***"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Imports"
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [],
"source": [
"import copy\n",
"import pandas as pd\n",
"import numpy as np\n",
"from scipy.stats import pointbiserialr, shapiro\n",
"\n",
"import matplotlib.pyplot as plt\n",
"import seaborn as sns"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Loading the All Clear dataset"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [],
"source": [
"properties = pd.read_csv(\"data/all_clear/lookup_properties.csv\", delimiter=';')\n",
"properties.set_index('#id', inplace=True)\n",
"rankings = pd.read_csv(\"data/all_clear/rankings.csv\", delimiter=';')\n",
"rankings.set_index('image_id', inplace=True)\n",
"rankings.sort_index(inplace=True)"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>filename</th>\n",
" <th>zooniverse_id</th>\n",
" <th>angle</th>\n",
" <th>area</th>\n",
" <th>areafrac</th>\n",
" <th>areathesh</th>\n",
" <th>bipolesep</th>\n",
" <th>c1flr24hr</th>\n",
" <th>id_filename</th>\n",
" <th>flux</th>\n",
" <th>...</th>\n",
" <th>hcpos_x</th>\n",
" <th>hcpos_y</th>\n",
" <th>m1flr12hr</th>\n",
" <th>m5flr12hr</th>\n",
" <th>n_nar</th>\n",
" <th>noaa</th>\n",
" <th>pxpos_x</th>\n",
" <th>pxpos_y</th>\n",
" <th>sszn</th>\n",
" <th>zurich</th>\n",
" </tr>\n",
" <tr>\n",
" <th>#id</th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>1</th>\n",
" <td>530be1183ae74079c300000d.jpg</td>\n",
" <td>ASZ000090y</td>\n",
" <td>37.8021</td>\n",
" <td>34400.0</td>\n",
" <td>0.12</td>\n",
" <td>2890.0</td>\n",
" <td>3.72</td>\n",
" <td>0</td>\n",
" <td>1</td>\n",
" <td>2.180000e+22</td>\n",
" <td>...</td>\n",
" <td>452.26991</td>\n",
" <td>443.92976</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>1</td>\n",
" <td>8809</td>\n",
" <td>229.19344</td>\n",
" <td>166.87700</td>\n",
" <td>1</td>\n",
" <td>bxo</td>\n",
" </tr>\n",
" <tr>\n",
" <th>2</th>\n",
" <td>530be1183ae74079c300000f.jpg</td>\n",
" <td>ASZ000090o</td>\n",
" <td>37.3590</td>\n",
" <td>78700.0</td>\n",
" <td>-0.00</td>\n",
" <td>6170.0</td>\n",
" <td>7.28</td>\n",
" <td>0</td>\n",
" <td>2</td>\n",
" <td>5.760000e+22</td>\n",
" <td>...</td>\n",
" <td>149.64301</td>\n",
" <td>621.53865</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>1</td>\n",
" <td>8810</td>\n",
" <td>200.41511</td>\n",
" <td>154.54088</td>\n",
" <td>2</td>\n",
" <td>fao</td>\n",
" </tr>\n",
" <tr>\n",
" <th>3</th>\n",
" <td>530be1183ae74079c3000011.jpg</td>\n",
" <td>ASZ0000946</td>\n",
" <td>58.6197</td>\n",
" <td>37900.0</td>\n",
" <td>0.08</td>\n",
" <td>937.0</td>\n",
" <td>3.88</td>\n",
" <td>0</td>\n",
" <td>3</td>\n",
" <td>2.150000e+22</td>\n",
" <td>...</td>\n",
" <td>704.04967</td>\n",
" <td>-436.33152</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>1</td>\n",
" <td>8812</td>\n",
" <td>205.30165</td>\n",
" <td>154.98689</td>\n",
" <td>3</td>\n",
" <td>axx</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4</th>\n",
" <td>530be1183ae74079c3000013.jpg</td>\n",
" <td>ASZ000090v</td>\n",
" <td>32.3099</td>\n",
" <td>31200.0</td>\n",
" <td>0.12</td>\n",
" <td>1720.0</td>\n",
" <td>4.90</td>\n",
" <td>0</td>\n",
" <td>4</td>\n",
" <td>1.660000e+22</td>\n",
" <td>...</td>\n",
" <td>-449.47446</td>\n",
" <td>-234.01929</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>1</td>\n",
" <td>8813</td>\n",
" <td>207.95782</td>\n",
" <td>169.12196</td>\n",
" <td>4</td>\n",
" <td>dro</td>\n",
" </tr>\n",
" <tr>\n",
" <th>5</th>\n",
" <td>530be1183ae74079c3000015.jpg</td>\n",
" <td>ASZ000090x</td>\n",
" <td>49.9221</td>\n",
" <td>88400.0</td>\n",
" <td>0.05</td>\n",
" <td>6480.0</td>\n",
" <td>12.48</td>\n",
" <td>0</td>\n",
" <td>5</td>\n",
" <td>6.130000e+22</td>\n",
" <td>...</td>\n",
" <td>-735.40990</td>\n",
" <td>208.46232</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>2</td>\n",
" <td>8814</td>\n",
" <td>183.10649</td>\n",
" <td>165.00398</td>\n",
" <td>5</td>\n",
" <td>hax</td>\n",
" </tr>\n",
" <tr>\n",
" <th>6</th>\n",
" <td>530be1183ae74079c3000017.jpg</td>\n",
" <td>ASZ000090q</td>\n",
" <td>41.7276</td>\n",
" <td>66500.0</td>\n",
" <td>-0.04</td>\n",
" <td>5450.0</td>\n",
" <td>7.45</td>\n",
" <td>0</td>\n",
" <td>6</td>\n",
" <td>4.510000e+22</td>\n",
" <td>...</td>\n",
" <td>307.16437</td>\n",
" <td>621.26440</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>1</td>\n",
" <td>8810</td>\n",
" <td>200.89942</td>\n",
" <td>158.90942</td>\n",
" <td>6</td>\n",
" <td>fao</td>\n",
" </tr>\n",
" <tr>\n",
" <th>7</th>\n",
" <td>530be1183ae74079c3000019.jpg</td>\n",
" <td>ASZ000090w</td>\n",
" <td>21.4421</td>\n",
" <td>31300.0</td>\n",
" <td>0.05</td>\n",
" <td>1930.0</td>\n",
" <td>5.32</td>\n",
" <td>1</td>\n",
" <td>7</td>\n",
" <td>1.580000e+22</td>\n",
" <td>...</td>\n",
" <td>-232.85591</td>\n",
" <td>-226.07368</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>1</td>\n",
" <td>8813</td>\n",
" <td>202.77628</td>\n",
" <td>166.14022</td>\n",
" <td>7</td>\n",
" <td>dso</td>\n",
" </tr>\n",
" <tr>\n",
" <th>8</th>\n",
" <td>530be1183ae74079c300001b.jpg</td>\n",
" <td>ASZ000090p</td>\n",
" <td>30.3312</td>\n",
" <td>49700.0</td>\n",
" <td>0.28</td>\n",
" <td>2670.0</td>\n",
" <td>1.86</td>\n",
" <td>0</td>\n",
" <td>8</td>\n",
" <td>2.430000e+22</td>\n",
" <td>...</td>\n",
" <td>-464.25581</td>\n",
" <td>230.57762</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>1</td>\n",
" <td>8814</td>\n",
" <td>191.95662</td>\n",
" <td>171.67525</td>\n",
" <td>8</td>\n",
" <td>hhx</td>\n",
" </tr>\n",
" <tr>\n",
" <th>9</th>\n",
" <td>530be1183ae74079c300001d.jpg</td>\n",
" <td>ASZ00008tg</td>\n",
" <td>42.8451</td>\n",
" <td>33600.0</td>\n",
" <td>-0.20</td>\n",
" <td>3850.0</td>\n",
" <td>6.03</td>\n",
" <td>1</td>\n",
" <td>9</td>\n",
" <td>2.840000e+22</td>\n",
" <td>...</td>\n",
" <td>-649.56237</td>\n",
" <td>212.13884</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>1</td>\n",
" <td>8815</td>\n",
" <td>197.90666</td>\n",
" <td>174.86663</td>\n",
" <td>9</td>\n",
" <td>cro</td>\n",
" </tr>\n",
" <tr>\n",
" <th>10</th>\n",
" <td>530be1183ae74079c300001f.jpg</td>\n",
" <td>ASZ00008sz</td>\n",
" <td>48.1351</td>\n",
" <td>51300.0</td>\n",
" <td>0.06</td>\n",
" <td>4750.0</td>\n",
" <td>7.52</td>\n",
" <td>1</td>\n",
" <td>10</td>\n",
" <td>4.060000e+22</td>\n",
" <td>...</td>\n",
" <td>452.20395</td>\n",
" <td>616.94779</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>1</td>\n",
" <td>8810</td>\n",
" <td>203.09814</td>\n",
" <td>158.22423</td>\n",
" <td>10</td>\n",
" <td>eao</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"<p>10 rows × 22 columns</p>\n",
"</div>"
],
"text/plain": [
" filename zooniverse_id angle area areafrac \\\n",
"#id \n",
"1 530be1183ae74079c300000d.jpg ASZ000090y 37.8021 34400.0 0.12 \n",
"2 530be1183ae74079c300000f.jpg ASZ000090o 37.3590 78700.0 -0.00 \n",
"3 530be1183ae74079c3000011.jpg ASZ0000946 58.6197 37900.0 0.08 \n",
"4 530be1183ae74079c3000013.jpg ASZ000090v 32.3099 31200.0 0.12 \n",
"5 530be1183ae74079c3000015.jpg ASZ000090x 49.9221 88400.0 0.05 \n",
"6 530be1183ae74079c3000017.jpg ASZ000090q 41.7276 66500.0 -0.04 \n",
"7 530be1183ae74079c3000019.jpg ASZ000090w 21.4421 31300.0 0.05 \n",
"8 530be1183ae74079c300001b.jpg ASZ000090p 30.3312 49700.0 0.28 \n",
"9 530be1183ae74079c300001d.jpg ASZ00008tg 42.8451 33600.0 -0.20 \n",
"10 530be1183ae74079c300001f.jpg ASZ00008sz 48.1351 51300.0 0.06 \n",
"\n",
" areathesh bipolesep c1flr24hr id_filename flux ... \\\n",
"#id ... \n",
"1 2890.0 3.72 0 1 2.180000e+22 ... \n",
"2 6170.0 7.28 0 2 5.760000e+22 ... \n",
"3 937.0 3.88 0 3 2.150000e+22 ... \n",
"4 1720.0 4.90 0 4 1.660000e+22 ... \n",
"5 6480.0 12.48 0 5 6.130000e+22 ... \n",
"6 5450.0 7.45 0 6 4.510000e+22 ... \n",
"7 1930.0 5.32 1 7 1.580000e+22 ... \n",
"8 2670.0 1.86 0 8 2.430000e+22 ... \n",
"9 3850.0 6.03 1 9 2.840000e+22 ... \n",
"10 4750.0 7.52 1 10 4.060000e+22 ... \n",
"\n",
" hcpos_x hcpos_y m1flr12hr m5flr12hr n_nar noaa pxpos_x \\\n",
"#id \n",
"1 452.26991 443.92976 0 0 1 8809 229.19344 \n",
"2 149.64301 621.53865 0 0 1 8810 200.41511 \n",
"3 704.04967 -436.33152 0 0 1 8812 205.30165 \n",
"4 -449.47446 -234.01929 0 0 1 8813 207.95782 \n",
"5 -735.40990 208.46232 0 0 2 8814 183.10649 \n",
"6 307.16437 621.26440 0 0 1 8810 200.89942 \n",
"7 -232.85591 -226.07368 0 0 1 8813 202.77628 \n",
"8 -464.25581 230.57762 0 0 1 8814 191.95662 \n",
"9 -649.56237 212.13884 0 0 1 8815 197.90666 \n",
"10 452.20395 616.94779 0 0 1 8810 203.09814 \n",
"\n",
" pxpos_y sszn zurich \n",
"#id \n",
"1 166.87700 1 bxo \n",
"2 154.54088 2 fao \n",
"3 154.98689 3 axx \n",
"4 169.12196 4 dro \n",
"5 165.00398 5 hax \n",
"6 158.90942 6 fao \n",
"7 166.14022 7 dso \n",
"8 171.67525 8 hhx \n",
"9 174.86663 9 cro \n",
"10 158.22423 10 eao \n",
"\n",
"[10 rows x 22 columns]"
]
},
"execution_count": 3,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"properties.head(10)"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>#id</th>\n",
" <th>count</th>\n",
" <th>k_value</th>\n",
" <th>score</th>\n",
" <th>std_dev</th>\n",
" </tr>\n",
" <tr>\n",
" <th>image_id</th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>1</th>\n",
" <td>2779</td>\n",
" <td>50</td>\n",
" <td>8</td>\n",
" <td>1126.778324</td>\n",
" <td>1.707604</td>\n",
" </tr>\n",
" <tr>\n",
" <th>2</th>\n",
" <td>10010</td>\n",
" <td>50</td>\n",
" <td>8</td>\n",
" <td>1312.434736</td>\n",
" <td>2.397493</td>\n",
" </tr>\n",
" <tr>\n",
" <th>3</th>\n",
" <td>10</td>\n",
" <td>50</td>\n",
" <td>8</td>\n",
" <td>962.280235</td>\n",
" <td>1.945574</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4</th>\n",
" <td>5718</td>\n",
" <td>50</td>\n",
" <td>8</td>\n",
" <td>1199.623395</td>\n",
" <td>1.894883</td>\n",
" </tr>\n",
" <tr>\n",
" <th>5</th>\n",
" <td>8599</td>\n",
" <td>50</td>\n",
" <td>8</td>\n",
" <td>1270.270911</td>\n",
" <td>1.207558</td>\n",
" </tr>\n",
" <tr>\n",
" <th>6</th>\n",
" <td>9385</td>\n",
" <td>50</td>\n",
" <td>8</td>\n",
" <td>1293.377418</td>\n",
" <td>1.454489</td>\n",
" </tr>\n",
" <tr>\n",
" <th>7</th>\n",
" <td>7288</td>\n",
" <td>50</td>\n",
" <td>8</td>\n",
" <td>1236.575624</td>\n",
" <td>2.318417</td>\n",
" </tr>\n",
" <tr>\n",
" <th>8</th>\n",
" <td>6152</td>\n",
" <td>50</td>\n",
" <td>8</td>\n",
" <td>1209.593634</td>\n",
" <td>1.790289</td>\n",
" </tr>\n",
" <tr>\n",
" <th>9</th>\n",
" <td>9607</td>\n",
" <td>50</td>\n",
" <td>8</td>\n",
" <td>1299.871029</td>\n",
" <td>3.015982</td>\n",
" </tr>\n",
" <tr>\n",
" <th>10</th>\n",
" <td>9828</td>\n",
" <td>50</td>\n",
" <td>8</td>\n",
" <td>1306.593595</td>\n",
" <td>1.143226</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" #id count k_value score std_dev\n",
"image_id \n",
"1 2779 50 8 1126.778324 1.707604\n",
"2 10010 50 8 1312.434736 2.397493\n",
"3 10 50 8 962.280235 1.945574\n",
"4 5718 50 8 1199.623395 1.894883\n",
"5 8599 50 8 1270.270911 1.207558\n",
"6 9385 50 8 1293.377418 1.454489\n",
"7 7288 50 8 1236.575624 2.318417\n",
"8 6152 50 8 1209.593634 1.790289\n",
"9 9607 50 8 1299.871029 3.015982\n",
"10 9828 50 8 1306.593595 1.143226"
]
},
"execution_count": 4,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"rankings.head(10)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Data Preparation\n",
"Making a new dataframe that holds information for any flare being produced and the associated AR complexity."
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {},
"outputs": [],
"source": [
"observed_columns = ['c1flr24hr', 'm1flr12hr', 'm5flr12hr', 'noaa']\n",
"flares_complexities = copy.deepcopy(properties[observed_columns])\n",
"flares_complexities['complexity'] = rankings.score"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>c1flr24hr</th>\n",
" <th>m1flr12hr</th>\n",
" <th>m5flr12hr</th>\n",
" <th>noaa</th>\n",
" <th>complexity</th>\n",
" </tr>\n",
" <tr>\n",
" <th>#id</th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>1</th>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>8809</td>\n",
" <td>1126.778324</td>\n",
" </tr>\n",
" <tr>\n",
" <th>2</th>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>8810</td>\n",
" <td>1312.434736</td>\n",
" </tr>\n",
" <tr>\n",
" <th>3</th>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>8812</td>\n",
" <td>962.280235</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4</th>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>8813</td>\n",
" <td>1199.623395</td>\n",
" </tr>\n",
" <tr>\n",
" <th>5</th>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>8814</td>\n",
" <td>1270.270911</td>\n",
" </tr>\n",
" <tr>\n",
" <th>6</th>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>8810</td>\n",
" <td>1293.377418</td>\n",
" </tr>\n",
" <tr>\n",
" <th>7</th>\n",
" <td>1</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>8813</td>\n",
" <td>1236.575624</td>\n",
" </tr>\n",
" <tr>\n",
" <th>8</th>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>8814</td>\n",
" <td>1209.593634</td>\n",
" </tr>\n",
" <tr>\n",
" <th>9</th>\n",
" <td>1</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>8815</td>\n",
" <td>1299.871029</td>\n",
" </tr>\n",
" <tr>\n",
" <th>10</th>\n",
" <td>1</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>8810</td>\n",
" <td>1306.593595</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" c1flr24hr m1flr12hr m5flr12hr noaa complexity\n",
"#id \n",
"1 0 0 0 8809 1126.778324\n",
"2 0 0 0 8810 1312.434736\n",
"3 0 0 0 8812 962.280235\n",
"4 0 0 0 8813 1199.623395\n",
"5 0 0 0 8814 1270.270911\n",
"6 0 0 0 8810 1293.377418\n",
"7 1 0 0 8813 1236.575624\n",
"8 0 0 0 8814 1209.593634\n",
"9 1 0 0 8815 1299.871029\n",
"10 1 0 0 8810 1306.593595"
]
},
"execution_count": 6,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"flares_complexities.head(10)"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {},
"outputs": [],
"source": [
"def combine_flaring_columns(ar):\n",
" ar['flares'] = ar.m5flr12hr | ar.m1flr12hr | ar.c1flr24hr\n",
" ar.drop(['m5flr12hr', 'm1flr12hr', 'c1flr24hr'], axis=1, inplace=True)"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {},
"outputs": [],
"source": [
"combine_flaring_columns(flares_complexities)"
]
},
{
"cell_type": "code",
"execution_count": 9,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>noaa</th>\n",
" <th>complexity</th>\n",
" <th>flares</th>\n",
" </tr>\n",
" <tr>\n",
" <th>#id</th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>1</th>\n",
" <td>8809</td>\n",
" <td>1126.778324</td>\n",
" <td>0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>2</th>\n",
" <td>8810</td>\n",
" <td>1312.434736</td>\n",
" <td>0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>3</th>\n",
" <td>8812</td>\n",
" <td>962.280235</td>\n",
" <td>0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4</th>\n",
" <td>8813</td>\n",
" <td>1199.623395</td>\n",
" <td>0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>5</th>\n",
" <td>8814</td>\n",
" <td>1270.270911</td>\n",
" <td>0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>6</th>\n",
" <td>8810</td>\n",
" <td>1293.377418</td>\n",
" <td>0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>7</th>\n",
" <td>8813</td>\n",
" <td>1236.575624</td>\n",
" <td>1</td>\n",
" </tr>\n",
" <tr>\n",
" <th>8</th>\n",
" <td>8814</td>\n",
" <td>1209.593634</td>\n",
" <td>0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>9</th>\n",
" <td>8815</td>\n",
" <td>1299.871029</td>\n",
" <td>1</td>\n",
" </tr>\n",
" <tr>\n",
" <th>10</th>\n",
" <td>8810</td>\n",
" <td>1306.593595</td>\n",
" <td>1</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" noaa complexity flares\n",
"#id \n",
"1 8809 1126.778324 0\n",
"2 8810 1312.434736 0\n",
"3 8812 962.280235 0\n",
"4 8813 1199.623395 0\n",
"5 8814 1270.270911 0\n",
"6 8810 1293.377418 0\n",
"7 8813 1236.575624 1\n",
"8 8814 1209.593634 0\n",
"9 8815 1299.871029 1\n",
"10 8810 1306.593595 1"
]
},
"execution_count": 9,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"flares_complexities.head(10)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Seperating the positive and negative class\n"
]
},
{
"cell_type": "code",
"execution_count": 10,
"metadata": {},
"outputs": [],
"source": [
"does_flare = flares_complexities[flares_complexities.flares == 1]\n",
"does_not_flare = flares_complexities[flares_complexities.flares == 0]"
]
},
{
"cell_type": "code",
"execution_count": 11,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>noaa</th>\n",
" <th>complexity</th>\n",
" <th>flares</th>\n",
" </tr>\n",
" <tr>\n",
" <th>#id</th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>7</th>\n",
" <td>8813</td>\n",
" <td>1236.575624</td>\n",
" <td>1</td>\n",
" </tr>\n",
" <tr>\n",
" <th>9</th>\n",
" <td>8815</td>\n",
" <td>1299.871029</td>\n",
" <td>1</td>\n",
" </tr>\n",
" <tr>\n",
" <th>10</th>\n",
" <td>8810</td>\n",
" <td>1306.593595</td>\n",
" <td>1</td>\n",
" </tr>\n",
" <tr>\n",
" <th>14</th>\n",
" <td>8810</td>\n",
" <td>1300.302886</td>\n",
" <td>1</td>\n",
" </tr>\n",
" <tr>\n",
" <th>23</th>\n",
" <td>8816</td>\n",
" <td>1180.945863</td>\n",
" <td>1</td>\n",
" </tr>\n",
" <tr>\n",
" <th>32</th>\n",
" <td>8814</td>\n",
" <td>1369.772259</td>\n",
" <td>1</td>\n",
" </tr>\n",
" <tr>\n",
" <th>40</th>\n",
" <td>8829</td>\n",
" <td>1099.494215</td>\n",
" <td>1</td>\n",
" </tr>\n",
" <tr>\n",
" <th>41</th>\n",
" <td>8824</td>\n",
" <td>1349.701272</td>\n",
" <td>1</td>\n",
" </tr>\n",
" <tr>\n",
" <th>48</th>\n",
" <td>8829</td>\n",
" <td>1353.643280</td>\n",
" <td>1</td>\n",
" </tr>\n",
" <tr>\n",
" <th>51</th>\n",
" <td>8824</td>\n",
" <td>1414.873927</td>\n",
" <td>1</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" noaa complexity flares\n",
"#id \n",
"7 8813 1236.575624 1\n",
"9 8815 1299.871029 1\n",
"10 8810 1306.593595 1\n",
"14 8810 1300.302886 1\n",
"23 8816 1180.945863 1\n",
"32 8814 1369.772259 1\n",
"40 8829 1099.494215 1\n",
"41 8824 1349.701272 1\n",
"48 8829 1353.643280 1\n",
"51 8824 1414.873927 1"
]
},
"execution_count": 11,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"does_flare.head(10)"
]
},
{
"cell_type": "code",
"execution_count": 12,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>noaa</th>\n",
" <th>complexity</th>\n",
" <th>flares</th>\n",
" </tr>\n",
" <tr>\n",
" <th>#id</th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>1</th>\n",
" <td>8809</td>\n",
" <td>1126.778324</td>\n",
" <td>0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>2</th>\n",
" <td>8810</td>\n",
" <td>1312.434736</td>\n",
" <td>0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>3</th>\n",
" <td>8812</td>\n",
" <td>962.280235</td>\n",
" <td>0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4</th>\n",
" <td>8813</td>\n",
" <td>1199.623395</td>\n",
" <td>0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>5</th>\n",
" <td>8814</td>\n",
" <td>1270.270911</td>\n",
" <td>0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>6</th>\n",
" <td>8810</td>\n",
" <td>1293.377418</td>\n",
" <td>0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>8</th>\n",
" <td>8814</td>\n",
" <td>1209.593634</td>\n",
" <td>0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>11</th>\n",
" <td>8813</td>\n",
" <td>1243.681834</td>\n",
" <td>0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>12</th>\n",
" <td>8814</td>\n",
" <td>1191.450319</td>\n",
" <td>0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>13</th>\n",
" <td>8815</td>\n",
" <td>1218.807609</td>\n",
" <td>0</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" noaa complexity flares\n",
"#id \n",
"1 8809 1126.778324 0\n",
"2 8810 1312.434736 0\n",
"3 8812 962.280235 0\n",
"4 8813 1199.623395 0\n",
"5 8814 1270.270911 0\n",
"6 8810 1293.377418 0\n",
"8 8814 1209.593634 0\n",
"11 8813 1243.681834 0\n",
"12 8814 1191.450319 0\n",
"13 8815 1218.807609 0"
]
},
"execution_count": 12,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"does_not_flare.head(10)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"***"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Point-Biserial Correlation\n",
"A point-biserial correlation is used to measure the strength and direction of the association that exists between one continuous variable and one dichotomous variable. It is a special case of the Pearson’s product-moment correlation, which is applied when you have two continuous variables, whereas in this case one of the variables is measured on a dichotomous scale.\n",
"\n",
"## Assumptions for using Point-Biserial Correlation\n",
"\n",
"* **Assumption 1**: One of the two variables should be measured on a continuous scale. In this analysis, the `Complexity` is continous.\n",
"\n",
"* **Assumption 2**: The other variable should be dichotomous. In this analysis, the whether an AR `flares` is dichotomous, **_0_** denoting no flaring and **_1_** denoting flaring.\n",
"\n",
"* **Assumption 3**: The continuous variable should have equal variances for each category of the dichotomous variable.\n",
"\n",
"* **Assumption 4**: There should be no outliers for the continuous variable for each category of the dichotomous variable.\n",
"\n",
"* **Assumption 5**: The continuous variable should be approximately normally distributed for each category of the dichotomous variable.\n",
"\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"***"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### For Assumption 3"
]
},
{
"cell_type": "code",
"execution_count": 13,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"8871.866873500114"
]
},
"execution_count": 13,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"does_flare.complexity.var()"
]
},
{
"cell_type": "code",
"execution_count": 14,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"10177.684638071736"
]
},
"execution_count": 14,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"does_not_flare.complexity.var()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### This violates assumption 3.\n",
"To fix it, we normalize the complexities in each class."
]
},
{
"cell_type": "code",
"execution_count": 15,
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"/home/apollo/anaconda3/envs/Andromeda/lib/python3.6/site-packages/ipykernel_launcher.py:1: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" \"\"\"Entry point for launching an IPython kernel.\n",
"/home/apollo/anaconda3/envs/Andromeda/lib/python3.6/site-packages/ipykernel_launcher.py:2: SettingWithCopyWarning: \n",
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
"Try using .loc[row_indexer,col_indexer] = value instead\n",
"\n",
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
" \n"
]
}
],
"source": [
"does_flare['normalized_complexity'] = (does_flare.complexity - does_flare.complexity.mean()) / does_flare.complexity.std()\n",
"does_not_flare['normalized_complexity'] = (does_not_flare.complexity - does_not_flare.complexity.mean()) / does_not_flare.complexity.std()"
]
},
{
"cell_type": "code",
"execution_count": 16,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>noaa</th>\n",
" <th>complexity</th>\n",
" <th>flares</th>\n",
" <th>normalized_complexity</th>\n",
" </tr>\n",
" <tr>\n",
" <th>#id</th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>7</th>\n",
" <td>8813</td>\n",
" <td>1236.575624</td>\n",
" <td>1</td>\n",
" <td>-0.927213</td>\n",
" </tr>\n",
" <tr>\n",
" <th>9</th>\n",
" <td>8815</td>\n",
" <td>1299.871029</td>\n",
" <td>1</td>\n",
" <td>-0.255221</td>\n",
" </tr>\n",
" <tr>\n",
" <th>10</th>\n",
" <td>8810</td>\n",
" <td>1306.593595</td>\n",
" <td>1</td>\n",
" <td>-0.183849</td>\n",
" </tr>\n",
" <tr>\n",
" <th>14</th>\n",
" <td>8810</td>\n",
" <td>1300.302886</td>\n",
" <td>1</td>\n",
" <td>-0.250636</td>\n",
" </tr>\n",
" <tr>\n",
" <th>23</th>\n",
" <td>8816</td>\n",
" <td>1180.945863</td>\n",
" <td>1</td>\n",
" <td>-1.517822</td>\n",
" </tr>\n",
" <tr>\n",
" <th>32</th>\n",
" <td>8814</td>\n",
" <td>1369.772259</td>\n",
" <td>1</td>\n",
" <td>0.486905</td>\n",
" </tr>\n",
" <tr>\n",
" <th>40</th>\n",
" <td>8829</td>\n",
" <td>1099.494215</td>\n",
" <td>1</td>\n",
" <td>-2.382575</td>\n",
" </tr>\n",
" <tr>\n",
" <th>41</th>\n",
" <td>8824</td>\n",
" <td>1349.701272</td>\n",
" <td>1</td>\n",
" <td>0.273816</td>\n",
" </tr>\n",
" <tr>\n",
" <th>48</th>\n",
" <td>8829</td>\n",
" <td>1353.643280</td>\n",
" <td>1</td>\n",
" <td>0.315667</td>\n",
" </tr>\n",
" <tr>\n",
" <th>51</th>\n",
" <td>8824</td>\n",
" <td>1414.873927</td>\n",
" <td>1</td>\n",
" <td>0.965739</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" noaa complexity flares normalized_complexity\n",
"#id \n",
"7 8813 1236.575624 1 -0.927213\n",
"9 8815 1299.871029 1 -0.255221\n",
"10 8810 1306.593595 1 -0.183849\n",
"14 8810 1300.302886 1 -0.250636\n",
"23 8816 1180.945863 1 -1.517822\n",
"32 8814 1369.772259 1 0.486905\n",
"40 8829 1099.494215 1 -2.382575\n",
"41 8824 1349.701272 1 0.273816\n",
"48 8829 1353.643280 1 0.315667\n",
"51 8824 1414.873927 1 0.965739"
]
},
"execution_count": 16,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"does_flare.head(10)"
]
},
{
"cell_type": "code",
"execution_count": 17,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>noaa</th>\n",
" <th>complexity</th>\n",
" <th>flares</th>\n",
" <th>normalized_complexity</th>\n",
" </tr>\n",
" <tr>\n",
" <th>#id</th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>1</th>\n",
" <td>8809</td>\n",
" <td>1126.778324</td>\n",
" <td>0</td>\n",
" <td>-0.680140</td>\n",
" </tr>\n",
" <tr>\n",
" <th>2</th>\n",
" <td>8810</td>\n",
" <td>1312.434736</td>\n",
" <td>0</td>\n",
" <td>1.160147</td>\n",
" </tr>\n",
" <tr>\n",
" <th>3</th>\n",
" <td>8812</td>\n",
" <td>962.280235</td>\n",
" <td>0</td>\n",
" <td>-2.310698</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4</th>\n",
" <td>8813</td>\n",
" <td>1199.623395</td>\n",
" <td>0</td>\n",
" <td>0.041924</td>\n",
" </tr>\n",
" <tr>\n",
" <th>5</th>\n",
" <td>8814</td>\n",
" <td>1270.270911</td>\n",
" <td>0</td>\n",
" <td>0.742205</td>\n",
" </tr>\n",
" <tr>\n",
" <th>6</th>\n",
" <td>8810</td>\n",
" <td>1293.377418</td>\n",
" <td>0</td>\n",
" <td>0.971244</td>\n",
" </tr>\n",
" <tr>\n",
" <th>8</th>\n",
" <td>8814</td>\n",
" <td>1209.593634</td>\n",
" <td>0</td>\n",
" <td>0.140752</td>\n",
" </tr>\n",
" <tr>\n",
" <th>11</th>\n",
" <td>8813</td>\n",
" <td>1243.681834</td>\n",
" <td>0</td>\n",
" <td>0.478646</td>\n",
" </tr>\n",
" <tr>\n",
" <th>12</th>\n",
" <td>8814</td>\n",
" <td>1191.450319</td>\n",
" <td>0</td>\n",
" <td>-0.039090</td>\n",
" </tr>\n",
" <tr>\n",
" <th>13</th>\n",
" <td>8815</td>\n",
" <td>1218.807609</td>\n",
" <td>0</td>\n",
" <td>0.232084</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" noaa complexity flares normalized_complexity\n",
"#id \n",
"1 8809 1126.778324 0 -0.680140\n",
"2 8810 1312.434736 0 1.160147\n",
"3 8812 962.280235 0 -2.310698\n",
"4 8813 1199.623395 0 0.041924\n",
"5 8814 1270.270911 0 0.742205\n",
"6 8810 1293.377418 0 0.971244\n",
"8 8814 1209.593634 0 0.140752\n",
"11 8813 1243.681834 0 0.478646\n",
"12 8814 1191.450319 0 -0.039090\n",
"13 8815 1218.807609 0 0.232084"
]
},
"execution_count": 17,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"does_not_flare.head(10)"
]
},
{
"cell_type": "code",
"execution_count": 18,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"1.0"
]
},
"execution_count": 18,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"does_not_flare.normalized_complexity.var()"
]
},
{
"cell_type": "code",
"execution_count": 19,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"0.9999999999999967"
]
},
"execution_count": 19,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"does_flare.normalized_complexity.var()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### This assumption 3 holds.\n",
"Continous variable in both positive and negative has equal variance."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"****"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Combining the positive and negative classes"
]
},
{
"cell_type": "code",
"execution_count": 20,
"metadata": {},
"outputs": [],
"source": [
"all_flares = pd.concat([does_flare, does_not_flare])"
]
},
{
"cell_type": "code",
"execution_count": 21,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>noaa</th>\n",
" <th>complexity</th>\n",
" <th>flares</th>\n",
" <th>normalized_complexity</th>\n",
" </tr>\n",
" <tr>\n",
" <th>#id</th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>7326</th>\n",
" <td>9965</td>\n",
" <td>1233.908404</td>\n",
" <td>0</td>\n",
" <td>0.381768</td>\n",
" </tr>\n",
" <tr>\n",
" <th>6722</th>\n",
" <td>9868</td>\n",
" <td>1242.323301</td>\n",
" <td>0</td>\n",
" <td>0.465179</td>\n",
" </tr>\n",
" <tr>\n",
" <th>12464</th>\n",
" <td>10773</td>\n",
" <td>1281.222179</td>\n",
" <td>0</td>\n",
" <td>0.850758</td>\n",
" </tr>\n",
" <tr>\n",
" <th>6002</th>\n",
" <td>9744</td>\n",
" <td>1121.378700</td>\n",
" <td>0</td>\n",
" <td>-0.733663</td>\n",
" </tr>\n",
" <tr>\n",
" <th>6400</th>\n",
" <td>9799</td>\n",
" <td>1279.962526</td>\n",
" <td>0</td>\n",
" <td>0.838272</td>\n",
" </tr>\n",
" <tr>\n",
" <th>11038</th>\n",
" <td>10554</td>\n",
" <td>1363.359946</td>\n",
" <td>0</td>\n",
" <td>1.664934</td>\n",
" </tr>\n",
" <tr>\n",
" <th>1576</th>\n",
" <td>9054</td>\n",
" <td>1279.053561</td>\n",
" <td>0</td>\n",
" <td>0.829262</td>\n",
" </tr>\n",
" <tr>\n",
" <th>10426</th>\n",
" <td>10454</td>\n",
" <td>1137.855032</td>\n",
" <td>0</td>\n",
" <td>-0.570344</td>\n",
" </tr>\n",
" <tr>\n",
" <th>11102</th>\n",
" <td>10564</td>\n",
" <td>1333.140412</td>\n",
" <td>1</td>\n",
" <td>0.097993</td>\n",
" </tr>\n",
" <tr>\n",
" <th>12626</th>\n",
" <td>10794</td>\n",
" <td>1171.999114</td>\n",
" <td>1</td>\n",
" <td>-1.612808</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" noaa complexity flares normalized_complexity\n",
"#id \n",
"7326 9965 1233.908404 0 0.381768\n",
"6722 9868 1242.323301 0 0.465179\n",
"12464 10773 1281.222179 0 0.850758\n",
"6002 9744 1121.378700 0 -0.733663\n",
"6400 9799 1279.962526 0 0.838272\n",
"11038 10554 1363.359946 0 1.664934\n",
"1576 9054 1279.053561 0 0.829262\n",
"10426 10454 1137.855032 0 -0.570344\n",
"11102 10564 1333.140412 1 0.097993\n",
"12626 10794 1171.999114 1 -1.612808"
]
},
"execution_count": 21,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"all_flares.sample(frac=1).head(10)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"***"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### For Assumption 4"
]
},
{
"cell_type": "code",
"execution_count": 22,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 1200x1200 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"plt.figure(figsize=(12,12), dpi=100)\n",
"sns.boxplot(y=all_flares.normalized_complexity, x=all_flares.flares)\n",
"plt.show()"
]
},
{
"cell_type": "code",
"execution_count": 23,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 1200x1200 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"plt.figure(figsize=(12,12), dpi=100)\n",
"sns.violinplot(y=all_flares.normalized_complexity, x=all_flares.flares)\n",
"plt.show()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### There are some outliers for the positive class (Need help to deal with them)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"***"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### For Assumption 5"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Plotting the distribution of positive and negative classes"
]
},
{
"cell_type": "code",
"execution_count": 24,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 1200x1200 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"plt.figure(figsize=(12,12), dpi=100)\n",
"sns.distplot(does_flare.normalized_complexity, color='green')\n",
"sns.distplot(does_not_flare.normalized_complexity, color='red')\n",
"plt.legend(labels=['Flare occurs', 'Flare does not occur'])\n",
"plt.show()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### To test if distrbution can be assumed gaussian"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Shapiro-Wilks test for normality\n",
"The Shapiro-Wilks test for normality is used to detect all departures from normality.\n",
"\n",
"The test rejects the hypothesis of normality when the p-value is less than or equal to 0.05. \n",
"\n",
"Passing the normality test only indicates that no significant departure from normality was found."
]
},
{
"cell_type": "code",
"execution_count": 25,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"The p value for this test is 1.3814005237797574e-26\n"
]
}
],
"source": [
"statistic, p_value = shapiro(does_flare.complexity)\n",
"print(f\"The p value for this test is {p_value}\")"
]
},
{
"cell_type": "code",
"execution_count": 26,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"The p value for this test is 8.128411264416912e-19\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"/home/apollo/anaconda3/envs/Andromeda/lib/python3.6/site-packages/scipy/stats/morestats.py:1676: UserWarning: p-value may not be accurate for N > 5000.\n",
" warnings.warn(\"p-value may not be accurate for N > 5000.\")\n"
]
}
],
"source": [
"statistic, p_value = shapiro(does_not_flare.complexity)\n",
"print(f\"The p value for this test is {p_value}\")"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Both distributions fail the Shapiro test. Need help to fix this."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"***"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Point-Biserial Correlation\n"
]
},
{
"cell_type": "code",
"execution_count": 27,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>noaa</th>\n",
" <th>complexity</th>\n",
" <th>flares</th>\n",
" <th>normalized_complexity</th>\n",
" </tr>\n",
" <tr>\n",
" <th>#id</th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>7728</th>\n",
" <td>10028</td>\n",
" <td>1119.386192</td>\n",
" <td>0</td>\n",
" <td>-0.753413</td>\n",
" </tr>\n",
" <tr>\n",
" <th>7182</th>\n",
" <td>9933</td>\n",
" <td>1120.718174</td>\n",
" <td>0</td>\n",
" <td>-0.740210</td>\n",
" </tr>\n",
" <tr>\n",
" <th>12101</th>\n",
" <td>10723</td>\n",
" <td>1139.305594</td>\n",
" <td>0</td>\n",
" <td>-0.555966</td>\n",
" </tr>\n",
" <tr>\n",
" <th>6023</th>\n",
" <td>9745</td>\n",
" <td>1151.497316</td>\n",
" <td>0</td>\n",
" <td>-0.435117</td>\n",
" </tr>\n",
" <tr>\n",
" <th>10321</th>\n",
" <td>10432</td>\n",
" <td>1078.626608</td>\n",
" <td>0</td>\n",
" <td>-1.157435</td>\n",
" </tr>\n",
" <tr>\n",
" <th>829</th>\n",
" <td>8943</td>\n",
" <td>1234.441695</td>\n",
" <td>0</td>\n",
" <td>0.387054</td>\n",
" </tr>\n",
" <tr>\n",
" <th>6193</th>\n",
" <td>9765</td>\n",
" <td>1258.035947</td>\n",
" <td>0</td>\n",
" <td>0.620928</td>\n",
" </tr>\n",
" <tr>\n",
" <th>302</th>\n",
" <td>8851</td>\n",
" <td>1112.909907</td>\n",
" <td>0</td>\n",
" <td>-0.817608</td>\n",
" </tr>\n",
" <tr>\n",
" <th>10300</th>\n",
" <td>10429</td>\n",
" <td>1174.858872</td>\n",
" <td>0</td>\n",
" <td>-0.203550</td>\n",
" </tr>\n",
" <tr>\n",
" <th>8266</th>\n",
" <td>10110</td>\n",
" <td>1234.766115</td>\n",
" <td>0</td>\n",
" <td>0.390270</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" noaa complexity flares normalized_complexity\n",
"#id \n",
"7728 10028 1119.386192 0 -0.753413\n",
"7182 9933 1120.718174 0 -0.740210\n",
"12101 10723 1139.305594 0 -0.555966\n",
"6023 9745 1151.497316 0 -0.435117\n",
"10321 10432 1078.626608 0 -1.157435\n",
"829 8943 1234.441695 0 0.387054\n",
"6193 9765 1258.035947 0 0.620928\n",
"302 8851 1112.909907 0 -0.817608\n",
"10300 10429 1174.858872 0 -0.203550\n",
"8266 10110 1234.766115 0 0.390270"
]
},
"execution_count": 27,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"all_flares.sample(frac=1).head(10)"
]
},
{
"cell_type": "code",
"execution_count": 28,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"PointbiserialrResult(correlation=0.4596014988642637, pvalue=0.0)"
]
},
"execution_count": 28,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"pointbiserialr(all_flares.flares, all_flares.complexity)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### The Point-Biserial Correlation shows a moderate positive correlation between AR complexity and Flare Production."
]
}
],
"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": "3.6.10"
}
},
"nbformat": 4,
"nbformat_minor": 4
}
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment