Basic table operations

This section will cover essential basic functionality in crandas, including inspecting, assigning new columns, computing with columns, boolean logic, and manipulating column names.

Inspecting CDataFrames

While you cannot normally access the data stored in a CDataFrame, you can use the CDataFrame.describe() method to inspect a CDataFrame to get summary statistics of the numeric data.

import crandas as cd

df = cd.DataFrame({'A': [1, 2, 3, 4, 5], 'B': [-1, -1, -1, -1, -1]})

>>> print(df.describe())
                    A        B
    0   type   integer  integer
    1  count         5        5
    2   mean       3.0     -1.0
    3    std  1.581139      0.0
    4    min         1       -1
    5    max         5       -1

See Working with numeric data for details

Assigning new columns

Assigning new columns to an existing table unfortunately is not as simple as it would be in pandas. Using df["new_col"] = [5,4,3,2,1] will not work in crandas. To add a column to a CDataFrame, we use the CDataFrame.assign() method.

df = cd.DataFrame({"col1":[1,2,3,4,5], "col2": [6,7,8,9,10]})

# Create a new column called 'col3' which is equal to 'col1' - 1
df = df.assign(col3=df["col1"] - 1)

# Create a 2 new columns: "col4" which is equal to 'col1' + 1, and "col5" which is equal to "col2" - 1
df = df.assign(col4=lambda x: x.col1 + 1, col5=lambda x: x.col2 - 1)

>>> print(df.open())
        col1  col2  col3  col4  col5
    0     1     6     0     2     5
    1     2     7     1     3     6
    2     3     8     2     4     7
    3     4     9     3     5     8
    4     5    10     4     6     9

Computing with columns

Given a column, we can check whether its values are equal, greater (or equal) than or less (or equal) than a given value.

df = cd.DataFrame({
   "id": [1,2,3,4,5],
    "is_member": [0,1,1,0,1],
    "height": [175,162,151,160,180]
    }
)

#We check who has id=2
id_is2 = df["id"] == 2

>>> print(id_is2.as_table().open())
    0  0
    1  1
    2  0
    3  0
    4  0

#We check which rows hold the information of someone who is 160cm or taller
at_least160 = df["height"] >= 160

>>> print(at_least160.as_table().open())
    0  1
    1  1
    2  0
    3  1
    4  1

Of course, we probably want to do more than just creating this column. We can also filter based on this value (Find out more about filtering in Selecting data):

df = cd.DataFrame({
   "id": [1,2,3,4,5],
    "is_member": [0,1,1,0,1],
    "height": [175,162,151,160,180]
    }
)

# Filter the df such that only members are included (using ==)
member = df[df["is_member"] == 1]

# Print the mean member height
print(member['height'].mean())

# For not_member we can change '==' to '!='
not_member = df[df['is_member'] != 1]

# Print the mean not_member height
print(not_member['height'].mean())

Note

You can also use boolean operations over columns, like (df['is_member'] > 3) & (df['height'] < 160). We support and &, or | and xor ^.

Conditionals over columns

Once you are able to check whether a column fulfills a certain property, you might want to create a new column with a value based on that. crandas has the CSeries.if_else() method to make this happen. For example, this can allow you to make categorical columns out of numerical ones, by nesting the function.

df = cd.DataFrame({
   "id": [1,2,3,4,5],
    "is_member": [0,1,1,0,1],
    "height": [175,162,151,160,180]
    }
)

# create a column that shows the cms above 160 in the person's height or -1 if they are shorter
df = df.assign(cms_above_160=(df["height"] >= 160).if_else(df["height"] - 160, -1))

# creates 3 categories for height: 0 for below 160, 1 for between 160 and 170 and 2 for above 170
df = df.assign(height_cats=(df["height"] >= 170).if_else(2, (df["height"] >= 160).if_else(1,0)))

Note

Crandas provides easier ways to create categorical columns, as seen in Categorical data.

Manipulating column names

Crandas provides methods to rename, add suffixes, or add prefixes to column names.

The CDataFrame.rename() allows you to rename specific columns by providing a dictionary, where the keys are the current column names, and the values are the new column names.

df = cd.DataFrame({"col1": [1,2,3], "col2": [4,5,6]})
df = df.rename(columns={"col1": "A", "col2": "B"})
>>> print(list(df.columns))
['A','B']

If instead you would like to add a prefix or a suffix to all column names in a CDataFrame then you should use either of the following methods:

• CDataFrame.add_suffix()

• CDataFrame.add_prefix()

df = cd.DataFrame({"col1": [1,2,3], "col2": [4,5,6]})

df = df.add_prefix("_1")

df = df.add_suffix('_2')

>>> print(list(df.columns))
['_1col1_2','_1col2_2']

Hint

Adding prefixes or suffixes to your column names can be useful when merging tables which have columns with the same names.

Now that we know how to work with columns, in the next section we will find out how to select specific rows and how to filter based on values found in the data.