NavData

This tutorial shows how to use the NavData class, including how to initialize instances and perform basic operations.

import numpy as np
import pandas as pd

import gnss_lib_py as glp

# Get data path of example file
glp.make_dir("../data")
!wget https://raw.githubusercontent.com/Stanford-NavLab/gnss_lib_py/main/notebooks/tutorials/data/myreceiver.csv --quiet -nc -O "../data/myreceiver.csv"
data_path = "../data/myreceiver.csv"

Initializing

Create an empty NavData class instance

empty_nav_data = glp.NavData()
print(empty_nav_data)

Empty NavData
Rows: []
Index: []

Create a NavData class from a csv file

nav_data_csv = glp.NavData(csv_path=data_path)
print(nav_data_csv)

   myTimestamp  mySatId  myPseudorange
0           10       10      270000001
1           10       14      270000007
2           10        7      270000004
3           10        3      270000005
4           11       10      270000002
5           11       14      270000008
6           11        7      270000003
7           11        3      270000004

Create an empty NavData class from a pandas DataFrame

# Convert the csv file to a pd.DataFrame for this illustration
pd_df = pd.read_csv(data_path)
# Load pd.DataFrame into NavData using pd.DataFrame instance
nav_data_pd = glp.NavData(pandas_df=pd_df)
print(nav_data_pd)

   myTimestamp  mySatId  myPseudorange
0           10       10      270000001
1           10       14      270000007
2           10        7      270000004
3           10        3      270000005
4           11       10      270000002
5           11       14      270000008
6           11        7      270000003
7           11        3      270000004

Create a NavData class instance from a numpy array

np_array = np.eye(4)
nav_data_np = glp.NavData(numpy_array=np_array)
print(nav_data_np)

     0    1    2    3
0  1.0  0.0  0.0  0.0
1  0.0  1.0  0.0  0.0
2  0.0  0.0  1.0  0.0
3  0.0  0.0  0.0  1.0

Accessing values

For this example, we will use the csv file that we previously loaded. This file has the following contents:

myTimestamp	mySatId	myPseudorange
10	G10	270000001
10	G14	270000007
10	G07	270000004
10	G03	270000005
11	G10	270000002
11	G14	270000008
11	G07	270000003
11	G03	270000004

NavData stores the different types of measurements, like myTimestamp as rows. So, each row contains the variable at different times/measurements. In this case, the myPseudorange row contains values for different timestamps and satellite IDs

Values in rows can be accessed by using the name of the row or the number of the row

nav_data_csv['myTimestamp']

array([10, 10, 10, 10, 11, 11, 11, 11])

nav_data_csv[0]

array([10, 10, 10, 10, 11, 11, 11, 11])

By specifying a column number (or slice of columns) along with the row label/number, a sub part of the matrix can be accessed as well

nav_data_csv['myPseudorange', :3]

array([270000001, 270000007, 270000004])

To access multiple rows using labels, pass the labels as a list of strings or slice of numbers

nav_data_csv[['myTimestamp', 'myPseudorange']]

array([[1.00000000e+01, 1.00000000e+01, 1.00000000e+01, 1.00000000e+01,
        1.10000000e+01, 1.10000000e+01, 1.10000000e+01, 1.10000000e+01],
       [2.70000001e+08, 2.70000007e+08, 2.70000004e+08, 2.70000005e+08,
        2.70000002e+08, 2.70000008e+08, 2.70000003e+08, 2.70000004e+08]])

However, to ensure compatability with accessing operations, you cannot access string and numeric values jointly. This is to prevent unexpected behaviour that might result on performing mathematical operations with strings and numbers.

If you attempt to access a combination of strings and numbers, the method will return an AssertionError

try:
    nav_data_csv[0:2]
except AssertionError as err:
    print(err)

Setting values

You can update values for existing rows by indexing those rows and assigning an array

nav_data_csv['myPseudorange'] = 10*np.arange(8)
nav_data_csv['myPseudorange']

array([ 0, 10, 20, 30, 40, 50, 60, 70])

You can also set individual values by accessing the specific rows and columns that you want to reassign

nav_data_csv['myPseudorange', 2] = 111
nav_data_csv['myPseudorange']

array([  0,  10, 111,  30,  40,  50,  60,  70])

Setting new row

To set numeric values for a new row, access the row with the label name and assign the value to that row.

nav_data_csv['new_row'] = np.arange(8)
nav_data_csv['new_row']

array([0, 1, 2, 3, 4, 5, 6, 7])

New string rows can also be created similarly.

Note that while creating rows with string values, you must set the dtype of the row to object

nav_data_csv['new_string_row'] = np.array(['string1', 'string1', 'string1', 'string1',
                                            'string2', 'string2', 'string2', 'string2'], dtype=object)
nav_data_csv['new_string_row']

array([np.str_('string1'), np.str_('string1'), np.str_('string1'),
       np.str_('string1'), np.str_('string2'), np.str_('string2'),
       np.str_('string2'), np.str_('string2')], dtype=object)

Find rows, shape and length

You can see which rows are currently in NavData using NavData.rows. The number of columns can be viewed using len(NavData) or NavData.num_cols and the shape of the array can be viewed using NavData.shape.

nav_data_csv.rows

['myTimestamp', 'mySatId', 'myPseudorange', 'new_row', 'new_string_row']

nav_data_csv.shape

(5, 8)

len(nav_data_csv)

8

nav_data_csv.num_cols

8

Removing rows or columns

To remove rows, use the NavData.remove() method, specifying the rows and columns you want to remove.

nav_data_removed = nav_data_csv.remove(rows=['new_row', 'new_string_row'])
print(nav_data_removed.rows)
print(nav_data_removed.shape)

['myTimestamp', 'mySatId', 'myPseudorange']
(3, 8)

nav_data_removed = nav_data_csv.remove(cols=[0, 1])
print(nav_data_removed.shape)

(5, 6)

Copy

For NavData.copy(), similar to the remove method, specify the rows and columns you want to copy into a new NavData instance.

Specifying rows, columns or both copies specified rows and columns.

Note that currently, rows and columns must be specified as lists.

nav_data_np_copy = nav_data_np.copy(rows=[0,1], cols=list(range(4)))
print('Original NavData array')
print(nav_data_np[:, :])
print('Copied NavData array')
print(nav_data_np_copy[:, :])

Original NavData array
[[1. 0. 0. 0.]
 [0. 1. 0. 0.]
 [0. 0. 1. 0.]
 [0. 0. 0. 1.]]
Copied NavData array
[[1. 0. 0. 0.]
 [0. 1. 0. 0.]]

If you don’t specify any rows or columns, the entire NavData will be copied

nav_data_np_copy = nav_data_np.copy()
print('Original NavData array')
print(nav_data_np[:, :])
print('Copied NavData array')
print(nav_data_np_copy[:, :])

Original NavData array
[[1. 0. 0. 0.]
 [0. 1. 0. 0.]
 [0. 0. 1. 0.]
 [0. 0. 0. 1.]]
Copied NavData array
[[1. 0. 0. 0.]
 [0. 1. 0. 0.]
 [0. 0. 1. 0.]
 [0. 0. 0. 1.]]

Convert to pd.DataFrame and save to csv

You can use the NavData.to_csv() method to save the data as a csv file and NavData.pandas_df() to get the equivalent pd.DataFrame

nav_data_pd.pandas_df()

	myTimestamp	mySatId	myPseudorange
0	10	10	270000001
1	10	14	270000007
2	10	7	270000004
3	10	3	270000005
4	11	10	270000002
5	11	14	270000008
6	11	7	270000003
7	11	3	270000004

To save, specify the path in NavData.to_csv()

# nav_data_pd.to_csv('desired path for csv file')

Looping over columns

You can loop over all columns using in-built looping functions. In this example, we terminate the loop after 3 runs for display clarity

for col_idx, nav_data_col in enumerate(nav_data_pd):
    print('Current column number', col_idx)
    print('Current column')
    print(nav_data_col)
    if col_idx >= 3:
        break

Current column number 0
Current column
   myTimestamp  mySatId  myPseudorange
0           10       10      270000001
Current column number 1
Current column
   myTimestamp  mySatId  myPseudorange
0           10       14      270000007
Current column number 2
Current column
   myTimestamp  mySatId  myPseudorange
0           10        7      270000004
Current column number 3
Current column
   myTimestamp  mySatId  myPseudorange
0           10        3      270000005