wasdi 0.6.5

The Wasdi Python library

WASDI Python Library

WASDI is the Web Advanced Space Developer Interface. This software is a preliminary version of the Python Library you can use to access the WASDI platform functionalities from your Python code.

Visit us at http://www.wasdi.net

The source code can be found here

---

Python tutorial

WASPY is the WASDI Python Library.

Prerequisites:

mandatory:

• a WASDI registered user (with a username/password, google users are not supported yet)
• at least one workspace
• some EO products in your workspace

Optional:

• SNAP Python (snappy) interface: this is not necessary but you may find these useful, especially for reading and writing images locally (howto). Anyway, most of the SNAP functionalities are wrapped by WASDI, so don't worry.

Installation

To start working with WASPY, just install the library using:

pip install wasdi

To quickly check if the installation worked correctly, try running the following code:

import wasdi
print(wasdi.hello())

You should see this kind of output:

{"boolValue":null,"doubleValue":null,"intValue":null,"stringValue":"Hello Wasdi!!"}

Configuration

Create a config.json file. It is a standard json file, which is used to store the credentials of the user and some other settings. The syntax is:

“VARIABLE_NAME”: value

Hint: exploit an editor which can check the syntax (there are many which can be accessed online for free)

The minimal configuration to begin working with WASPY is:

{
  "USER": "[e-mail]",
  "PASSWORD": "yourPasswordHere",
  "WORKSPACE": "nameOfTheWorkspaceYouWantToUse"
}

For the other available parameters please refer to the Documentation.

Start WASPY

To start WASPY and check if everything is working, run the following code:

wasdi.init('./config.json')

(Adapt the path if the file is not located in your working directory)

The Lib will read the configuration file, load the user and password, log the user in, and then open the workspace specified in the configuration file. To check if everything is working, try to get the list of workspaces available for the user:

wasdi.getWorkspaces()

You should be able to see a result similar to the following one:

[{u'ownerUserId': u'[e-mail]',
  u'sharedUsers': [],
  u'workspaceId': u'23ab54f3-b453-2b3e-284a-b6a4243f0f2c',
  u'workspaceName': u'nameOfTheWorkspaceYouWantToUse'},
 {u'ownerUserId': u'[e-mail]',
  u'sharedUsers': [],
  u'workspaceId': u'103fbf01-2e68-22d3-bd45-2cf95665dac2',
  u'workspaceName': u'theNameOfAnotherWorkspace'}]

The configured Workspace is already opened. The use can open another workspace using:

wasdi.openWorkspace('theNameOfAnotherWorkspace')

and the lib replies showing the workspace unique id:

u'9ce787d4-1d59-4146-8df7-3fc9516d4eb3'

To get the list of the products available in the workspace, call

wasdi.getProductsByWorkspace('nameOfTheWorkspaceYouWantToUse')

and the lib returns a list of the products in the given workspace:

[u'S1A_IW_GRDH_1SDV_20190517T053543_20190517T053608_027263_0312F1_F071.zip',
u'S1B_IW_RAW__0SDV_20190506T052631_20190506T052703_016119_01E53A_D2AD.zip', u'S1A_IW_GRDH_1SDV_20190517T053608_20190517T053633_027263_0312F1_3382.zip']

Now try something more, let's search for some Sentinel 1 images. Let's assume we are interested in images taken from "2018-09-01" to "2018-09-02". Also, we'd better specify a bounding box. Assume we're interested in images with latitude in [43, 44] and longitude in [11, 12]. We can think of these coordinates as a rectangle limited by the upper left corner (44, 11) and the lower right corner(43, 12). The corresponding code is:

wasdi.wasdiLog('Let\'s search some images')
aoImages = wasdi.searchEOImages("S1", "2018-09-01", "2018-09-02", 44, 11, 43, 12, None, None, None, None)
wasdi.wasdiLog('Found ' + str(len(aoImages)))

The output should be similar to this:

 Let's search some images
[INFO] waspy.searchEOImages: search results:
[{
		'footprint': 'POLYGON ((8.8724 45.3272, 8.4505 43.3746, 11.4656 43.0981, 11.9901 45.0472, 8.8724 45.3272, 8.8724 45.3272))',
		'id': 'cba6c104-3006-4af7-a2d1-cbd55f58b939',
		'link': 'https://catalogue.onda-dias.eu/dias-catalogue/Products(cba6c104-3006-4af7-a2d1-cbd55f58b939)/$value',
		'preview': None,
		'properties': {
			'offline': 'false',
			'downloadable': '',
			'filename': 'S1A_IW_RAW__0SDV_20180902T052727_20180902T052759_023515_028F75_7325.zip',
			'size': '1.54 GB',
			'pseudopath': 'RADAR/LEVEL-0/2018/09/02, S1/1A/SAR-C/LEVEL-0/IW_RAW__0S/2018/09/02, S1/1A/LEVEL-0/IW_RAW__0S/2018/09/02, S1/SAR-C/LEVEL-0/IW_RAW__0S/2018/09/02, S1/LEVEL-0/IW_RAW__0S/2018/09/02, 2014-016A/SAR-C/LEVEL-0/IW_RAW__0S/2018/09/02, 2014-016A/LEVEL-0/IW_RAW__0S/2018/09/02',
			'link': 'https://catalogue.onda-dias.eu/dias-catalogue/Products(cba6c104-3006-4af7-a2d1-cbd55f58b939)/$value',
			'format': 'application/zip',
			'creationDate': '2018-09-03T05:12:37.000Z'
		},
		'provider': 'ONDA',
		'summary': 'Date: 2018-09-03T05:12:37.000Z, Instrument: null, Mode: null, Satellite: null, Size: 1.54 GB',
		'title': 'S1A_IW_RAW__0SDV_20180902T052727_20180902T052759_023515_028F75_7325'
},
{'(...7 more results similar to this one, omitted for brevity)'}]
Found 8

Now we can import one of those products in WASDI: let's download the first one:

sImportWithDict = wasdi.importProduct(None, None, aoImages[0])

We can see a list of the products in the workspace as follows:

asProducts = wasdi.getProductsByActiveWorkspace()
wasdi.wasdiLog(asProducts)

The second line logs the list of products

Running an existing workflow

If you wish to run an existing SNAP workflow you can use wasdi.executeWorkflow. For example, if you wish to execute a workflow that calibrates and corrects the georeference of a Sentinel 1 image, you may use the workflow called LISTSinglePreproc in this way:

asProducts = wasdi.getProductsByActiveWorkspace()
sStatus = wasdi.executeWorkflow([asProducts[0]], ['lovelyOutput'], 'LISTSinglePreproc')

Here the first line gets the list of products and the second calls the workflow LISTSinglePreproc on the first product of the workspace and creates another product called lovelyOutput.

A more complete example

Now put everything back together. Create a file called myProcessor.py (follow the link to download the file) with the following content:

import wasdi


def run(parameters, processId):
    wasdi.wasdiLog('Here\'s the list of your workspaces:')
    aoWorkspaces = wasdi.getWorkspaces()
    wasdi.wasdiLog(aoWorkspaces)
    wasdi.wasdiLog('The ID of currently selected workspace is:')
    sActiveWorkspace = wasdi.getActiveWorkspaceId()
    wasdi.wasdiLog(sActiveWorkspace)

    wasdi.wasdiLog('Let\'s search some images...')
    aoImages = wasdi.searchEOImages("S1", "2018-09-01", "2018-09-02", 44, 11, 43, 12, sProductType='GRD')
    wasdi.wasdiLog('Found ' + str(len(aoImages)) + ' images')

    wasdi.wasdiLog('Download the first one passing the dictionary...')
    sImportWithDict = wasdi.importProduct(None, None, aoImages[0])
    wasdi.wasdiLog('Import with dict returned: ' + sImportWithDict)

    wasdi.wasdiLog('Now, these are the products in your workspace: ')
    asProducts = wasdi.getProductsByActiveWorkspace()
    wasdi.wasdiLog(asProducts)

    wasdi.wasdiLog('Let\'s run a workflow on the first image to rectify its georeference...')
    sStatus = wasdi.executeWorkflow([asProducts[0]], ['lovelyOutput'], 'LISTSinglePreproc')
    if sStatus == 'DONE':
        wasdi.wasdiLog('The product is now in your workspace, look at it on the website')

    wasdi.wasdiLog('It\'s over!')

def WasdiHelp():
    sHelp = "Wasdi Tutorial"
    return sHelp

Then create another file to start the processor. Let's call it tutorial.py (follow the link to download the file), with the following content:

import myProcessor
import wasdi

bInitResult = wasdi.init('config.json')
if bInitResult:
    myProcessor.run(wasdi.getParametersDict(), '')

Now, if you run tutorial.py, it will call myProcessor.py, which will go through the instructions we saw above. Pro tip: keep the browser open in wasdi.net (make sure you are logged in) and open the workspace you are using, to see the evolution of the script in real time.

Deploy your processor on WASDI

Finally, to deply our processor on WASDI, you need first to create a text file called pip.txt (follow the link to download the file) containg the packages we imported in myProcessor.py, one per line. Since we just imported wasdi, it should look like this:

wasdi

Now, create a zip file containing these two files:

• myProcessor.py
• pip.txt

You can now upload the zip file on wasdi.net from Edit -> Processor -> New WASDI App by giving it a name and completing the other details. You will need to do this just once. To run it, go to WADI Apps -> (select yours) -> no parameters are needed, so just enter {} and clic run.

---

More to include WASDI in a custom Processor

Let’s assume that the developer has his own EO Product Manipulation Code. At some point, the developer wishes to read his own input file, then make elaborations and finally save an output file.

Let’s imagine a pseudo-code like this.

# Input and output file name
filename = '~wasdiUser/EO/myfile.zip'
outputfilename = "~wasdiUser/EO/myoutput.tiff"

# Read the file
EOimage = multibandRead(filename, size, precision, offset, interleave, byteorder)

# Elaborate the image somehow
EOimage *= 2

# Save the output
imwrite(EOimage, outputfilename)

To port the code onto WASDI, the pseudo-code has to be integrated like this:

import wasdi
import os

filename = 'myfile.zip'
outputFileName = 'myoutput.tiff'

fullInputPath = wasdi.getFullProductPath(filename)

# Read the file
EOproduct = multibandRead(fullInputPath, size, precision, offset, interleave, byteorder)

# Elaborate the image
EOproduct *= 2

# Save the output
# Get The Path
outputPath = wasdi.getSavePath()
fullOutputPath = os.path.join(outputPath, outputFileName)

# Use the save path
imwrite(EOproduct, fullOutputPath)

# Ingest in WASDI
wasdi.addFileToWASDI(outputFileName)

We modified the code to start the library and then to receive from WASDI the paths to use.

The input files are supposed to be in the workspace. In order for this to happen, the user can go the wasdi web application, open the workspace, search the needed image and add it to the workspace.

The wasdi.getFullProductPath method has a double goal:

1. as the name suggests, it returns the local path to use back to the developer
2. if the code is running on the client PC, the Wasdi Lib will checks if the file is available locally: in case this checks fails, the lib will automatically download the file from the WASDI cloud to the local PC.
   To disable the auto download feature, is possible to add this parameter to the config.json file:\

"DOWNLOADACTIVE":0

The choice of a name for the output file is left to the user, WASPY just provides the folder to use (wasdi.GetSavePath). So to save the file we need to get the path and then concatenate the custom file name (fullOutputPath = os.path.join(outputPath, outputFileName)).

The last call, AddFileToWASDI, has the goal to add the product to the workspace. It takes in input only the file name, without the full path.

When used on the local PC, it will automatically upload the file after writing it on local file system. To inhibit this behavior, just add the following to the config.json:\

"UPLOADACTIVE":0

Use Custom parameters

Every processor usually has its own parameters. A typical example can be the name of a file in input, a threshold, the coordinates of an area of interest and so on. To let the developer work with her/his own parameters, WASPY implements an automatic file read.

Add this line to the configuration file config.json:

"PARAMETERSFILEPATH": "<path to a similar file for own parameters>"

e.g.

"PARAMETERSFILEPATH": "c:/temp/myparameters.txt"

Then create the same file in the right folder and fill it with all the needed parameters, using the same syntax used for config.json; e.g.:

"INPUTFILE": "S1A_imported_file.zip",
"THRESHOLD": 5,
"POINT": [44.2, 23.4]

The decision about how to encode these parameters is left to the developer. For WASDI these are all strings. In the example above, the developers may know that THRESHOLD is a number, and POINT is couple of coordinates that must to be splitted.

The only limit is that each parameter has to be written in one line.

In WASPY there are these three methods available:

• wasdi.getParameter(sKey): return the value of the sKey Parameter
• wasdi.addParameter(sKey, sValue): updates the value of a Parameter (ONLY in memory NOT in the file)
• wasdi.refreshParameters(): reads the parameter file from disk again

Let’s update the code above to use the parameters file. First of all create a parameter file and set the name and path in the config.json file. The file (i.e., parameters.json) might look like this:

{
  "INPUT_FILE": "S1A_imported_file.zip",
  "OUTPUT_FILE": "FloodedArea.tif"
}

Then modify the code to read the parameters without using hard-coded input:

import wasdi
import os

# The input file is supposed to be in the workspace
# Read the file from parameters
filename = wasdi.getParameter("INPUT_FILE")
outputfilename = wasdi.getParameter("OUTPUT_FILE")

fullInputPath = wasdi.getFullProductPath(filename)

# Read the file
EOproduct = multibandRead(fullInputPath, size, precision, offset, interleave, byteorder)

# Elaborate the image
EOproduct  *= 2

# Save the output
# Get The Path
outputPath = wasdi.getSavePath()
fullOutputPath = os.path.join(outputPath, outputFileName)

# Use the save path
imwrite(EOproduct, fullOutputPath)

# Ingest in WASDI
wasdi.addFileToWASDI(outputFileName)