Python Wrapper

For ease of development in Python, Aiimi provide a python wrapper. Docstrings provide detailed documentation in your IDE of choice, but a high-level overview is also provided here.

Simple Example

# First import the InsightMaker API
from insight_maker.api import InsightMaker


# Initialise an API instance, as a minimum you must provide the host server, or it will assume localhost

# The code tries to avoid requiring key or username entry here, but it is possible.
# By default im looks for the key in "ds.key", you can also provide a path to a .key file anywhere on your machine,
# or simply provide the key string - this is using the 'key' parameter.
# IM will use your local domain username, so rarely will you need to specify this, the only real case will be if you
# are using a service account or similar. This would be using the 'username' parameter.
im = InsightMaker(host="localhost")

print("")

# First get an idea of what datasets are available
datasets = im.datasets()
print("Datasets:")
for dataset in datasets:
    print("{} : {}".format(dataset.description(), len(dataset)))
print("")

# Let's choose the first one to play with
play_dataset = datasets[0]

# Now lets get some statistics for this dataset so we can understand it better
stats = im.stats(datasets = play_dataset)
# stats can be parsed as a DataFrame for ease of exploration
stats_df = stats.to_data_frame()
print(stats_df)
print("")

# Lets clean up our list by striping out any features which appear on less than half of our document set
stats_df = stats_df[stats_df["percentage"] >= 0.5]
print("Cleaned:")
print(stats_df)
print("")

# Now let's do a search for these features on InsightMaker
results = im.search("*", dataset = play_dataset, fields = stats_df.index, size = 50)
# Results is a DataSample object, which as a DataFrame as a property, we can print that
print(results.df)

# N.B. "*" was used as the search query to return all. You may also enter full lucene syntax here, a Elasticsearch
# query DSL dictionary or use one of the helper objects from insight_maker.api.query_builders (such as ExistsQuery,
# RegexQuery, RangeQuery and many more).

# The logical operators, &, | and ~ may be applied to query helper objects to build more complex expressions.