Using HDF5 with pandas

We can also read and write HDF5 files with pandas. To read HDF5 files with pandas, they must first be created with it. For example, let's use pandas to create a HDF5 file containing global power values:

import pandas as pd
import numpy as np
arr = np.loadtxt('temp.csv', skiprows=1, usecols=(2,3), delimiter=',')
import pandas as pd
store=pd.HDFStore('hdfstore_demo.hdf5')
print(store)
store['global_power']=pd.DataFrame(arr)
store.close()

Now let's read the HDF5 file that we created and print the array back:

import pandas as pd
store=pd.HDFStore('hdfstore_demo.hdf5')
print(store)
print(store['global_power'])
store.close()

The values of the DataFrame can be read in three different ways:

• store['global_power']
• store.get('global_power')
• store.global_power

pandas also provides the high-level read_hdf() function and the to_hdf() DataFrame method for reading and writing HDF5 files.

More documentation on HDF5 in pandas is available at the following link: http://pandas.pydata.org/pandas-docs/stable/io.html#io-hdf5.