Installation

    pip install pandas

    Usage

    Pandas uses NumPy for most data types. See full documentation.

    import pandas as pd

    Input and output

    Read and write files with different formats. See documentation.

    Read

    # csv
df = pd.read_csv(path)

# excel
df = pd.read_excel(path)

# custom delimiter
df = pd.read_csv(path, delimiter='\t')

    Write

    # csv
df.to_csv('path.csv', index=False)

# excel
# one sheet
df.to_excel("path.xlsx", sheet_name='Sheet_name_1')

# more than one
df2 = df1.copy()
with pd.ExcelWriter('output.xlsx') as writer:
    df1.to_excel(writer, sheet_name='Sheet_name_1')
    df2.to_excel(writer, sheet_name='Sheet_name_2')

    Accessing rows and columns

    Selecting and accessing different data. See user guide.

    # get the first 4 rows
df.head(4)

# get the last 4 rows
df.tail(4)

# get headers
df.columns

# read specific column
df['A']

# read multiple columns
df[['A', 'B']]

# read specific row
df[0]

# read multiple rows
df[0:4]

# read specific location (row=2, column=1)
df.iloc[2,1]

# loop through each row
for index, row in df.iterrows():
    print(index, row)

# filter
# will give every row with the type value fire
df.loc[df['A'] == 'test']

    Sorting and describing data

    # will give count, mean, standard deviation, min, max, 25%, 50%, 75%
df.describe()

# sort by column
df.sort_values('A', ascending=True)

# will sort first by name then by type
df.sort_values(['A', 'B'], ascending=True)

# will sort first by name then by type
# the first with ascending=True and the last with ascending=False
df.sort_values(['A', 'B'], ascending=[1, 0])

    Manipulate data

    # create a new column with the values from column 'name' and 'type'
# it makes sense if 'name' and 'type' are numbers
df['C'] = df['A'] + df['B']
# or
df['C'] = df.iloc[:, :].sum(axis=1)

# remove column
df = df.drop(columns=['C'])

    Plotting