A wrapper library for the novem.io data visualisation platform. Create charts, documents, e-mails, and dashboards through one simple API.

Note: novem is currently in closed alpha. If you want to try it out, please reach out to hello@novem.io

Create a linechart from a pandas dataframe (this assumes a configured profile, described under "Getting started" below).

import numpy as np
import pandas as pd
from novem import Plot
# a sample price-like series; swap in your own dataframe. Name the index and
# series so the CSV has a proper "Date,Price" header.
dates = pd.date_range("2015-01-01", "2021-12-31", freq="B", name="Date")
prices = pd.Series(
 100 + np.random.randn(len(dates)).cumsum(), index=dates, name="Price"
)
line = Plot("price_hist", type="line", name="Sample price history")
# send data to the plot
prices.pipe(line)
# url to view the plot
print(line.url)

To get started with novem you will have to register an account. Please reach out to us!

Once you have a username and password you can setup your environment using:

 python -m novem --init

In addition to invoking the novem module as shown above, the novem package also includes an extensive command-line interface (cli). Check out CLI.md in this repository or novem.io for more details.

Every novem object needs a token and an API root to talk to the platform. These are resolved, in order of precedence, from:

1. explicit keyword arguments on the object (or a Session, see below)
2. values set programmatically via novem.config
3. the NOVEM_TOKEN / NOVEM_API_ROOT environment variables
4. the config file written by python -m novem --init

The simplest setup is the config file (--init above). To configure novem programmatically instead (handy in notebooks, scripts, or CI), set a token on the global novem.config object once, and objects created afterwards pick it up automatically:

import novem
novem.config.set_token("your-token")
plot = novem.Plot("my-plot")

novem.config also exposes set_api_root(...) (to point at a non-default API) and use_profile(...) (to select a profile from the config file).

Alternatively, pass the token straight to the object. An explicit argument always wins over whatever is on novem.config:

plot = novem.Plot("my-plot", token="your-token")

A Session captures connection settings (token, api_root, or a config-file profile) and constructs objects bound to them without touching the global defaults — useful when you work against several accounts at once:

import novem
work = novem.Session(profile="work")
personal = novem.Session(profile="personal")
# copy a plot's data from one account to the other
personal.Plot("earnings").data = work.Plot("earnings").data

Novem represents plots as a Plot class that can be imported from the main novem package from novem import Plot.

The Plot class takes a single mandatory positional argument, the name of the plot.

• If the plot name is new, the instantiation of the class will create the plot.
• If the plot name already exists, then the new object will operate on the existing plot.

In addition to the name, there are two broad categories of options for a plot: data and config.

• The data contains the actual information to visualise (usually in the form of numeric csv)
• Config, which contains information about the visual such as:
  • Type (bar, line, donut, map etc)
  • Titles/captions/names/colors/legends/axis etc

There are two ways to interact with the plots: you can either supply all the necessary options as named arguments when creating the plot, or use the property accessors to modify them one by one (this is more helpful when working with the plot interactively). Below is an example of both approaches.

from novem import Plot
# everything in the constructor
barchart = Plot(<name>, \
 type="bar", \
 title="barchart title", \
 caption = "caption"
)
# property approach
barchart = Plot("plot_name")
barchart.type = "bar"
barchart.title = "barchart title"
barchart.caption = "caption"

In addition to setting individual properties, the Plot object is also callable. This means that the resulting plot can be used as a function, either by being provided data as an argument, or used as part of a pipe chain.

from novem import Plot
import pandas as pd
import numpy as np
# construct some random sample data
df = pd.DataFrame(np.random.randn(100, 4), columns=list("ABCD")).cumsum()
line = Plot("new_line", type="line")
# alternative one, setting data explicitly to a csv string
line.data = df.to_csv()
# or let the plot invoke the to_csv
line.data = df
# alternative two, calling Plot with a csv string
line(df.to_csv())
# alternative three calling the Plot with an object that has a to_csv function
line(df)
# or
df.pipe(line)

Note: All novem plot operations are live. This means that as soon as you write to or modify any aspects of the plot object, those changes are reflected on the novem server and anyone watching the plot in real time.

Every novem operation talks to the platform over the network, so any call can fail: a rejected token, a missing resource, or a value the server won't accept. When a read or write fails, novem raises a NovemException (or a more specific subclass) carrying the server's error message instead of failing silently.

from novem import Plot
from novem.exceptions import NovemException, Novem403
try:
 # the constructor creates the plot (a PUT) unless create=False, so this
 # call can raise too, not only the writes that follow it
 plot = Plot("my-plot")
 plot.type = "line"
except Novem403:
 print("not permitted to write this plot")
except NovemException as e:
 # the exception carries the server message and any rejected lines
 print(f"novem rejected the write: {e}")

Every exception is importable from novem.exceptions and inherits from NovemException, so catching the base class catches them all:

• NovemException: the base class, raised for any otherwise-unclassified API error.
• Novem401: the server rejected the supplied token.
• Novem403: the token is valid but lacks permission for this operation.
• Novem404: the resource does not exist.
• NovemAuthError: no usable token could be resolved locally. This is distinct from Novem401, which is the server rejecting a token that was sent.

Note: creating an object that already exists is not an error — the create PUT returns a 409, which novem treats as a no-op rather than raising.

novem requests time out instead of hanging indefinitely. Every call uses a default timeout of (10s connect, 2min read). Job execution (job.run()) can take far longer and uses (30s connect, 30min read). A request that exceeds its timeout raises requests.exceptions.Timeout.

The novem python library and platform is under active development, contributions or issues are most welcome.

For guidelines on how to contribute, please check out the CONTRIBUTING.md file in this repository.

This python library is licensed under the MIT license, see the LICENSE file for details.