The data consists of the magnitudes of galaxies measured with various filters, and their redshift, which was measured using spectroscopy. You will need to predict the redshift of galaxies based on only the photometric data.
The dataset was generated with SDDS and downloaded form SciServer. The file's are gzipped, you will able to read them using the 'read_csv' function in pandas.
File descriptions
train.csv.gz
- the training set
test.csv.gz
- the test set
sampleSubmission.csv
- a sample submission file in the correct format
Data fields
id
- an anonymous id unique to a given galaxy
u,g,r,i,z
- magnitudes of galaxies measured with different color filters.
redshift
- the redshift of the galaxies measured with spectroscopy