Skip to main content

[re]ctangular[d]ata[frames]

Project description

redframes
PyPI PyPI - Python Version Pandas Version

redframes (rectangular data frames) is a data manipulation library for ML and visualization. It is fully interoperable with pandas, compatible with scikit-learn, and works great with matplotlib!

redframes prioritizes syntax over flexibility and scope. And minimizes the number-of-googles-per-lines-of-code™ so that you can focus on the work that matters most.

"What is redframes?" would be the answer to the Jeopardy! clue "A pythonic dplyr".

Install & Import

pip install redframes
import redframes as rf

Quickstart

Copy-and-paste this:

import redframes as rf

df = rf.DataFrame({
    "foo": ["A", "A", "B", None, "B", "A", "A", "C"],
    "bar": [1, 4, 2, -4, 5, 6, 6, -2], 
    "baz": [0.99, None, 0.25, 0.75, 0.66, 0.47, 0.48, None]
})

df["foo"] 
# ['A', 'A', 'B', None, 'B', 'A', 'A', 'C']
df.columns 
# ['foo', 'bar', 'baz']
df.dimensions
# {'rows': 8, 'columns': 3}
df.empty
# False
df.types
# {'foo': object, 'bar': int, 'baz': float}

(
    df
    .mutate({"bar100": lambda row: row["bar"] * 100})
    .select(["foo", "baz", "bar100"])
    .filter(lambda row: 
        (row["foo"].isin(["A", "B"])) & (row["bar100"] > 0)
    )
    .denix("baz")
    .group("foo")
    .rollup({
        "bar_mean": ("bar100", rf.stat.mean), 
        "baz_sum": ("baz", rf.stat.sum)
    })
    .gather(["bar_mean", "baz_sum"])
    .sort("value")
)

IO

Save, load, and convert rf.DataFrame objects:

import redframes as rf
import pandas as pd

df = rf.DataFrame({"foo": [1, 2], "bar": ["A", "B"]})

# save/load
rf.save(df, "example.csv")
df = rf.load("example.csv")

# to/from pandas
pandf = rf.unwrap(df)
reddf = rf.wrap(pandf)

Verbs

There are 23 core "verbs" that make up rf.DataFrame objects. Each verb is pure, "chain-able", and has an analog in pandas/dplyr (see docstrings for more info/examples):

pandas dplyr
.accumulate cumsum mutate(... = cumsum(...))
.append concat bind_rows
.combine + unite
.dedupe drop_duplicates distinct
.denix dropna drop_na
.drop drop(..., axis=1) select(- ...)
.fill fillna fill, replace_na
.filter df[df[col] == condition] filter
.gather melt gather, pivot_longer
.group groupby group_by
.join merge *_join
.mutate apply, astype mutate
.rank rank("dense") dense_rank
.rename rename rename
.replace replace mutate(... = case_when(...))
.rollup agg summarize
.sample sample(n, frac) sample_n, sample_frac
.select select select
.shuffle sample(frac=1) sample_frac(..., 1)
.sort sort_values arrange
.split df[col].str.split() separate
.spread pivot_table spread, pivot_wider
.take head, tail slice_head, slice_tail

matplotlib

rf.DataFrame objects integrate seamlessly with matplotlib:

import redframes as rf
import matplotlib.pyplot as plt

df = rf.DataFrame({
    'position': ['TE', 'K', 'RB', 'WR', 'QB'],
    'avp': [116.98, 131.15, 180, 222.22, 272.91]
})

df = (
    df
    .mutate({"color": lambda row: row["position"] in ["WR", "RB"]})
    .replace({"color": {False: "orange", True: "red"}})
)

plt.barh(df["position"], df["avp"], color=df["color"]);
redframes

scikit-learn

rf.DataFrame objects are fully compatible with sklearn functions, estimators, and transformers:

import redframes as rf
from sklearn.model_selection import train_test_split
from sklearn.linear_model import LinearRegression

df = rf.DataFrame({
    "touchdowns": [15, 19, 5, 7, 9, 10, 12, 22, 16, 10],
    "age": [21, 22, 21, 24, 26, 28, 30, 35, 28, 21],
    "mvp": [1, 1, 0, 0, 0, 0, 0, 1, 0, 0]
})

target = "touchdowns"
y = df[target]
X = df.drop(target)
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=1)

model = LinearRegression()
model.fit(X_train, y_train)
model.score(X_test, y_test)
# 0.5083194901655527

print(X_train.take(1))
# rf.DataFrame({'age': [21], 'mvp': [0]})

X_new = rf.DataFrame({'age': [22], 'mvp': [1]})
model.predict(X_new)
# array([19.])

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

redframes-1.1.tar.gz (24.8 kB view details)

Uploaded Source

Built Distribution

redframes-1.1-py3-none-any.whl (35.4 kB view details)

Uploaded Python 3

File details

Details for the file redframes-1.1.tar.gz.

File metadata

  • Download URL: redframes-1.1.tar.gz
  • Upload date:
  • Size: 24.8 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/4.0.1 CPython/3.9.14

File hashes

Hashes for redframes-1.1.tar.gz
Algorithm Hash digest
SHA256 e374acb383f6b0ad5e9e21e6dbbd22a7e8bba2156e70623fd3b4073ab55ee870
MD5 897a8cf0feb3f86dd55fe013c6220476
BLAKE2b-256 6b71cfb3471becfa93aec11eeddc1283feadf980b1d12196e142511aec2e54ca

See more details on using hashes here.

File details

Details for the file redframes-1.1-py3-none-any.whl.

File metadata

  • Download URL: redframes-1.1-py3-none-any.whl
  • Upload date:
  • Size: 35.4 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/4.0.1 CPython/3.9.14

File hashes

Hashes for redframes-1.1-py3-none-any.whl
Algorithm Hash digest
SHA256 d621a702dba10c364b974fc6cd433ffe72025b106b500cf7b0aa30fac4eb0fe3
MD5 b72ed847f31225450c815b3a17868341
BLAKE2b-256 9f990f65a269202ada2487f4cf024954eca968df5ba7dba8f84b46c039a684c1

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page