A lot of this info is in the docs for Tables.
Open the file for reading, and get a reference to a table:
>>> with tables.open_file(path, mode='r') as f:
... table = f.root.some_columns.table
How many records are in the table?
>>> table.nrows
Search for a matching row (e.g. row matching the value 12345)
>>> table = f.root.some_table.table
>>> table.read_where('some_column==12345')
array([(12345, 3928, 364544)]
dtype=[('some_column', '<i8'), ('foo', '<i8'), ('bar', '<i8')])
>>> for row in table.where('some_column==12345')
... print(row['some_column'], row['foo'], row['bar'])
Extract a single column for a matching row (similar to above):
>>> arr = table.read_where('some_column=12345')
>>> bar = arr[0][2]
Find a row for a given key (column value), and update a column in that row of the table:
>>> arr = table.get_where_list('some_column==12345')
>>> key = arr[0]
>>> table.cols.bar[key] = 42
Construct a dataframe with all values
>>> arr = table.read()
>>> df = pd.DataFrame(arr)
Add an index to a column of the table:
>>> table.cols.some_column.create_index()
>>> table.reindex() # to rebuild the index...
>>> table.reindex_dirty() # rebuild indexes, if they are dirty.
How to add blosc compression? (See the docs on Filters)
>>> blosc_filter = Filters(complevel=9, complib='blosc')
>>> table = fileh.create_table( ... )
>>> table.cols.some_column.create_index(filters=blosc_filter)