Usage¶

1) Objective¶

The pyclient is a framework developed in python to retrieve data from a Braincube. These data are then used by data-scientists to perform all kinds of independent analysis of Braincube.

Here is the class diagram:

2) Import¶

In [ ]:

from braincube_connector import client, braincube

3) Configuration¶

Creating a config object allows to indicate to the Pyclient the domain of the braincube targeted by the data extraction (domain) as well as the API key created in the Braincube account parameters (api_key).

In [ ]:

config = {
    "domain": "mybraincube.com",
    "api_key": "<my_personal_access_token>",
}

The key can also be put in a config.json in the local directory or in the file ~/.braincube/config.json.

Here's how to create an api key:

The necessary scopes are: SSO_READ and BRAINCUBE

4) Recovery of instances¶

Retrieving the instances related to the domain specified in the config object

client.get_instance(config_dict=config)

In [ ]:

client.get_instance(config_file="config.json")

5) Braincube recovery¶

Recovery of the Braincube targeted by extraction

In [ ]:

bc = braincube.get_braincube("demo")

6) Memorybase Recovery¶

In [ ]:

all_memory_base = bc.get_memory_base_list()
#print(all_memory_base)

7) Recovery of the targeted MemoryBase¶

Retrieving the targeted MemoryBase using its id

In [ ]:

mb = bc.get_memory_base(20)
print(mb)

8) Recovery of variables stored in the MemoryBase¶

Retrieving the list of all the variables present in the MemoryBase

In [ ]:

var_list = mb.get_variable_list()
print(var_list[0])

9) Retrieving information from a variable¶

In [ ]:

var = var_list[0]
var_name = var.get_name()
var_bcid = var.get_bcid()

print('name : {} | bcid : {}'.format(var_name, var_bcid))

10) Retrieving data from a variable¶

A- Create a filter :¶

Create a filter allows you to filter the data you want to recover, these can be of 7 different types:

Equals to
Select a data when the variable is equal to...
```
{
  "EQUALS": [ "mb20/d2000002", 2.0]
}
```

Between
Select a data when the variable is between...

{
  "BETWEEN": [ "mb20/d2000003", -1, 1]
}

Lower than
Select a data when the variable is smaller than...
```
{
  "LESS": [ "mb20/d2000003", 10]
}
```

Greater than
Select a data when the variable is larger than...
```
{
  "GREAT": [ "mb20/d2000003", 10]
}
```
Not:
Select a data when the variable is does not include the next filter.
```
{
  "Not": [{"filter":...}]
}
```

And gate
Filters can be added using AND.

{
  "AND": [{"filter1":...}, {"filter2":...}]
}

Or gate:
Filters can be combined using OR.

{
  "OR": [{"filter1":...}, {"filter2":...}]
}

In [ ]:

my_filters = [{"BETWEEN": ["mb20/d2000002",0,1]},{"BETWEEN": ["mb20/d2000003", -1, 1]}]

B- In dictionary form :¶

In [ ]:

data = mb.get_data(["2000077"], filters=my_filter)
print(data)

C- In the form of DataFrame pandas :¶

We can use the dataframe parameter of get_data() being a boolean indicating if the desired return is a DataFrame or not.

In [ ]:

data = mb.get_data(["2000077"], filters=my_filters, dataframe=True)
print(data)

D- Changing the label of the data¶

The get_data() function proposes a parameter called label_type taking the value "name" or "bcid". This parameter allows you to change the label of the returned data.

In [ ]:

data = mb.get_data(["2000001", "2000034"], filters=my_filters, label_type="name", dataframe=True)
print(data)

11) Parameters¶

The parameters are available in the parameter class of the braincube_connector. They are used to customize certain values of the braincube_connector.

In [ ]:

from braincube_connector import parameters

Changing the pagination of a query

In [ ]:

parameters.set_parameter({"page_size": 10})

Parsing dates in python's Datetime format

In [ ]:

parameters.set_parameter({"parse_date": True})

The Braincube database stores 3 names (tag, standard, or local) for a variable. By default, standard is used but you can change it with parameters :

In [ ]:

parameters.set_parameter(({"VariableDescription_name_key": "tag"}))

In [ ]:

data = mb.get_data(["2000001", "2000034"], filters=my_filter, label_type="name", dataframe=True)
print(data)