nsp_lesson2

Introduction to Pandas

Ch2 Series

• 1D data stream
  • Form of np.array, iterable, dict, or scalar
  • Contains an index for each element
    • A.k.a axis labels
    • Indices follow the keys of the dict if given a dict when initialising
  • Contains a name

Ch3-6 DataFrame

• 2D data table
  • Consists of a collection of series
  • Columns are tied together by an index
  • Creation using a list of tuples / dictionary of lists
  • From csv
    • df = pd.read_cv(dtype=str, index_col=n)
• Index columns by df.iloc[] with index
• Index columns by df.iloc[] with "name"
  • Slicing can be applied
• Iteration
  • By KV pair .iteritems()
  • By index row pair .iterrows()
  • By tuples .itertuples()
• Filtering
  • .drop_duplicates(inplace=True)
    • inplace=True option mutates the table
  • Apply condition to []
• Missing data
  • .isna().sum() or .isnull().sum() to see missing data summary
  • .dropna(inplace=True) removes all rows with null
  • .dropna(axis=1) remove all columns with null
• Mutating
  • .apply() can apply lambda functions to all elements / columns