Analysis of NASA Kepler Exoplanet Data

Objective: This project performs an exploratory data analysis on a dataset of over 2,300 confirmed exoplanets discovered by the Kepler Space Telescope. The goal is to identify key trends in exoplanet characteristics, including their size, temperature, and relationship to their host stars, culminating in an analysis to identify potentially habitable candidates.

Skills Demonstrated: Python, Pandas, Matplotlib, Data Cleaning, Data Visualization

import pandas as pd
print("Success! Pandas is ready to go.")

df = pd.read_csv('exoplanets.csv')
print("\nSuccessfully loaded the exoplanets.csv file. Here are the first 5 rows:")
df.head()

Success! Pandas is ready to go.

Successfully loaded the exoplanets.csv file. Here are the first 5 rows:

	kepid	kepoi_name	kepler_name	koi_disposition	koi_pdisposition	koi_score	koi_fpflag_nt	koi_fpflag_ss	koi_fpflag_co	koi_fpflag_ec	...	koi_steff_err2	koi_slogg	koi_slogg_err1	koi_slogg_err2	koi_srad	koi_srad_err1	koi_srad_err2	ra	dec	koi_kepmag
0	10797460	K00752.01	Kepler-227 b	CONFIRMED	CANDIDATE	1.000	0	0	0	0	...	-81.0	4.467	0.064	-0.096	0.927	0.105	-0.061	291.93423	48.141651	15.347
1	10797460	K00752.02	Kepler-227 c	CONFIRMED	CANDIDATE	0.969	0	0	0	0	...	-81.0	4.467	0.064	-0.096	0.927	0.105	-0.061	291.93423	48.141651	15.347
2	10811496	K00753.01	NaN	CANDIDATE	CANDIDATE	0.000	0	0	0	0	...	-176.0	4.544	0.044	-0.176	0.868	0.233	-0.078	297.00482	48.134129	15.436
3	10848459	K00754.01	NaN	FALSE POSITIVE	FALSE POSITIVE	0.000	0	1	0	0	...	-174.0	4.564	0.053	-0.168	0.791	0.201	-0.067	285.53461	48.285210	15.597
4	10854555	K00755.01	Kepler-664 b	CONFIRMED	CANDIDATE	1.000	0	0	0	0	...	-211.0	4.438	0.070	-0.210	1.046	0.334	-0.133	288.75488	48.226200	15.509

5 rows × 49 columns

confirmed_planets_df = df[df['koi_disposition'] == 'CONFIRMED']

print(f"original number of entries: {len(df)}")
print(f"Number of confirmed planets: {len(confirmed_planets_df)}")

confirmed_planets_df.head()

original number of entries: 9564
Number of confirmed planets: 2357

	kepid	kepoi_name	kepler_name	koi_disposition	koi_pdisposition	koi_score	koi_fpflag_nt	koi_fpflag_ss	koi_fpflag_co	koi_fpflag_ec	...	koi_steff_err2	koi_slogg	koi_slogg_err1	koi_slogg_err2	koi_srad	koi_srad_err1	koi_srad_err2	ra	dec	koi_kepmag
0	10797460	K00752.01	Kepler-227 b	CONFIRMED	CANDIDATE	1.000	0	0	0	0	...	-81.0	4.467	0.064	-0.096	0.927	0.105	-0.061	291.93423	48.141651	15.347
1	10797460	K00752.02	Kepler-227 c	CONFIRMED	CANDIDATE	0.969	0	0	0	0	...	-81.0	4.467	0.064	-0.096	0.927	0.105	-0.061	291.93423	48.141651	15.347
4	10854555	K00755.01	Kepler-664 b	CONFIRMED	CANDIDATE	1.000	0	0	0	0	...	-211.0	4.438	0.070	-0.210	1.046	0.334	-0.133	288.75488	48.226200	15.509
5	10872983	K00756.01	Kepler-228 d	CONFIRMED	CANDIDATE	1.000	0	0	0	0	...	-232.0	4.486	0.054	-0.229	0.972	0.315	-0.105	296.28613	48.224670	15.714
6	10872983	K00756.02	Kepler-228 c	CONFIRMED	CANDIDATE	1.000	0	0	0	0	...	-232.0	4.486	0.054	-0.229	0.972	0.315	-0.105	296.28613	48.224670	15.714

5 rows × 49 columns

confirmed_planets_df.describe()

	kepid	koi_score	koi_fpflag_nt	koi_fpflag_ss	koi_fpflag_co	koi_fpflag_ec	koi_period	koi_period_err1	koi_period_err2	koi_time0bk	...	koi_steff_err2	koi_slogg	koi_slogg_err1	koi_slogg_err2	koi_srad	koi_srad_err1	koi_srad_err2	ra	dec	koi_kepmag
count	2.357000e+03	2345.000000	2357.000000	2357.000000	2357.0	2357.0	2357.000000	2.356000e+03	2.356000e+03	2357.000000	...	2343.000000	2356.000000	2356.000000	2356.000000	2356.000000	2356.000000	2356.000000	2357.00000	2357.000000	2357.000000
mean	8.097056e+06	0.962857	0.198133	0.005940	0.0	0.0	27.706884	2.224126e-04	-2.224126e-04	157.578189	...	-124.297909	4.411244	0.089866	-0.106458	1.065797	0.175705	-0.149744	290.95981	44.360843	14.337261
std	2.584650e+06	0.143475	9.577990	0.076857	0.0	0.0	54.492454	1.260899e-03	1.260899e-03	42.687043	...	52.132108	0.235392	0.079409	0.064892	0.638132	0.143657	0.224811	4.75516	3.569286	1.229623
min	7.574500e+05	0.000000	0.000000	0.000000	0.0	0.0	0.341842	2.700000e-08	-5.100000e-02	120.565925	...	-712.000000	2.410000	0.000000	-0.748000	0.118000	0.000000	-4.943000	280.20660	36.577381	8.224000
25%	6.106282e+06	0.992000	0.000000	0.000000	0.0	0.0	5.144365	1.157500e-05	-1.300000e-04	134.250610	...	-152.000000	4.290000	0.036000	-0.126000	0.808000	0.066000	-0.161000	287.25668	41.420158	13.659000
50%	8.352537e+06	1.000000	0.000000	0.000000	0.0	0.0	11.448658	3.900000e-05	-3.900000e-05	140.678930	...	-112.000000	4.455000	0.064000	-0.105000	0.968000	0.160000	-0.089500	291.14951	44.327518	14.591000
75%	1.016627e+07	1.000000	0.000000	0.000000	0.0	0.0	25.961901	1.300000e-04	-1.157500e-05	171.376107	...	-83.500000	4.557000	0.130000	-0.052000	1.199250	0.243250	-0.055000	294.78915	47.182861	15.259000
max	1.283487e+07	1.000000	465.000000	1.000000	0.0	0.0	1071.232624	5.100000e-02	-2.700000e-08	589.719750	...	0.000000	5.274000	0.765000	0.000000	19.530000	2.551000	0.000000	301.72076	52.149132	17.475000

8 rows × 44 columns

confirmed_planets_df.isnull().sum()

kepid                   0
kepoi_name              0
kepler_name             0
koi_disposition         0
koi_pdisposition        0
koi_score              12
koi_fpflag_nt           0
koi_fpflag_ss           0
koi_fpflag_co           0
koi_fpflag_ec           0
koi_period              0
koi_period_err1         1
koi_period_err2         1
koi_time0bk             0
koi_time0bk_err1        1
koi_time0bk_err2        1
koi_impact              1
koi_impact_err1         1
koi_impact_err2         1
koi_duration            0
koi_duration_err1       1
koi_duration_err2       1
koi_depth               1
koi_depth_err1          1
koi_depth_err2          1
koi_prad                1
koi_prad_err1           1
koi_prad_err2           1
koi_teq                 1
koi_teq_err1         2357
koi_teq_err2         2357
koi_insol               1
koi_insol_err1          1
koi_insol_err2          1
koi_model_snr           1
koi_tce_plnt_num        8
koi_tce_delivname       8
koi_steff               1
koi_steff_err1          1
koi_steff_err2         14
koi_slogg               1
koi_slogg_err1          1
koi_slogg_err2          1
koi_srad                1
koi_srad_err1           1
koi_srad_err2           1
ra                      0
dec                     0
koi_kepmag              0
dtype: int64

Question 1: What is the size distribution of exoplanet sizes in terms of Earth's radius?

#import the plotting library
import matplotlib.pyplot as plt

# crate a histogram of the radii of confirmed planets
plt.hist(confirmed_planets_df['koi_prad'], bins=300)

# Add labels to the plot
plt.title('Distribution of Exoplanet Radii')
plt.xlabel('Planet Radius (in Earth Radii)')
plt.ylabel('Number of Planets')

# display plot
plt.show()

Conclusion: The histogram provides a clear picture of exoplanet size distribution. The vast majority of exoplanets fall between the range of 1 and 4 times earth's radius with single outliers reaching close to 80 times earth's radius. It is clear that most of the observable exoplanets are Super-Earths or Mini-Neptunes.

Question 2: How does the orbital period relate to temperature?

plt.figure(figsize=(10,6))
plt.scatter(confirmed_planets_df['koi_period'], confirmed_planets_df['koi_teq'])
plt.title('Planet Temperature vs. Orbital Period')
plt.xlabel('Orbital Period (days)')
plt.ylabel('Equilibrium Temperature (Kelvin)')

# Changes the x-axis to a log scale so the dense clusters will be more obvious
plt.xscale('log')

plt.show()

Conclusion: It is clear from the scatter plot that there is an inverse relationship between orbital period and the equilibrium temperature of the exoplanets. As the orbital period increases, the temperature decreases. Faster orbital periods result in higher equilibrium periods. This is consistent with physical laws, as planets with longer periods are typically further from their sun and as such receive less energy.

Question 3: What is the temperature distribution of the planet-hosting stars?

# Creating a histogram that shows the temperature of our planets with a 
# vertical line indicating the temperature of our sun

sun_temp = 5778

plt.hist(confirmed_planets_df['koi_steff'], bins=100)
plt.axvline(sun_temp, color='red', linestyle='dashed', linewidth=2)
plt.title('Distribution of Host Star Temperatures')
plt.xlabel('Star Temperature (Kelvin)')
plt.ylabel('Number of Stars')
plt.show()

Conclusion: It seems the majority of stars fall in a semi-gaussian distribution with our sun being close to the top of the bell curve. More stars tend to be on the colder side than on the higher range of temperatures indicating that much higher temperatures might not be suitable for the formation of exoplanets. The distribution peaks in the 4500K to 6000K range, indicating that K-type and G-type stars are the most common exoplanet hosts in this dataset.

Final Analysis: Identifying Potentially Habitable Planets

To find potentially habitable candidates we need to filter our data based on whether or not the planet is terrestrial and lcoated in the goldilocks zone. To that end I analyze which planets meet the following criteria.

• The insolation of the planet is between 0.4 and 1.5 times Earth's

• It is a terrestrial planet which can be indicated by the planet having a radius less than that of 2 earth radii.

First, I will create a dataframe containing only these exoplanets, then, I will create a visual comparison of these planets to all confirmed exoplanets to establish how common potentially habitable planets are.

# define the habitable planets
is_habitable = (
    (confirmed_planets_df['koi_prad'] < 2) &
    (confirmed_planets_df['koi_insol'] > 0.4) &
    (confirmed_planets_df['koi_insol'] < 1.5)
)

# Create the new DataFrame of habitable planets
habitable_planets_df = confirmed_planets_df[is_habitable]

# print the number of candidates found
print(f"Number of potentially habitable exoplanet candidates found: {len(habitable_planets_df)}")

# creating scatterplot of all confirmed planets
plt.figure(figsize=(12,7))
plt.scatter(
    confirmed_planets_df['koi_steff'],
    confirmed_planets_df['koi_insol'],
    c='grey',
    label='All Confirmed Planets'
)

#creating a scatterplot of all potentially habitable planets
plt.scatter(
    habitable_planets_df['koi_steff'],
    habitable_planets_df['koi_insol'],
    c='blue',
    label='Potentially Habitable Candidates'
)

# Add labels and convert y-axis to a logarithmic scale for visibility

plt.title('Habitable Zone Candidates')
plt.axhline(1.5,color='green', linestyle='dashed', linewidth=2, label = 'Habitable Zone Boundary')
plt.axhline(0.4,color='green', linestyle='dashed', linewidth=2)
plt.xlabel('Host Star Temperature (Kelvin)')
plt.ylabel('Stellar Insolation (Compared to Earth)')
plt.yscale('log')
plt.legend()
plt.show()

Number of potentially habitable exoplanet candidates found: 19

Conclusion: This data shows the expected. potentially habitable planets are exceptionally uncommon in the universe. With only 19 total candidates found out of 2357 total confirmed exoplanets. The scatter plot reveals that this is because most of the exoplanets receive a much higher dose of energy from their host star than is necessary for habitablility. Host star temperature does not appear to be the deciding factor in what prevents a planet from being habitable. In fact we can see that life could exist from relatively cold stars to stars that are hotter than our sun. This analysis demonstrates that while potentially habitable worlds are rare, they are not exclusive to sun-like stars, which broadens our search for life in the galaxy.