v0.4.5

Breaking Changes

Previously, if you want to compare Edit distance between one column and a single string, you would do

df.select(
    pds.str_leven(pl.col("c"), "word")
)

But now you have to do

df.select(
    pds.str_leven(pl.col("c"), pl.lit("word"))
)

The previous behavior will now look for a column named "word" instead of using the word "word."

What's Changed

• refactored psi and add benford law by @abstractqqq in #158
• Skip some rows in KNN by @DGolubets in #157
• Added data_mask in knn_ptwise by @abstractqqq in #161
• Pretty Code Snippets by @s1lvester in #162
• Tests for entrophies by @abstractqqq in #163
• Woe encoding by @abstractqqq in #164
• Added Median Absolute Deviation and renamed str2 to string by @abstractqqq in #165

New Contributors

• @DGolubets made their first contribution in #157
• @s1lvester made their first contribution in #162

Full Changelog: v0.4.4...v0.4.5

v0.4.4

What's Changed

• added par option in convolve by @abstractqqq in #146
• More dia plots by @abstractqqq in #147
• fixed corr method bug by @abstractqqq in #148
• Add string pre-preprocessing code by @CangyuanLi in #150
• String Cleaning by @CangyuanLi in #152
• Pipeline by @abstractqqq in #155

New Contributors

• @CangyuanLi made their first contribution in #150

Full Changelog: v0.4.3...v0.4.4

v0.4.3

What's Changed

• Kendall tau correlation coefficient by @abstractqqq in #137
• added combined pds.corr expression by @abstractqqq in #139
• new pca plot and lstsq plot by @abstractqqq in #140
• fixed a bug in lstsq regarding group by by @abstractqqq in #143
• added null dist by @abstractqqq in #144
• Improve convolve perf, especially when filter size is small. added method option by @abstractqqq in #145

Full Changelog: v0.4.2...v0.4.3

v0.4.2

Highlights

In Diagnosis, you can now generate 2d principal component graphs. This can help you visualize higher dimensional data.

dia = DIA(df)
dia.plot_pc2(pl.all().exclude("species"), by = "species")

In addition, the following PCA related queries are available:

df.select(
    pds.query_pca("a", "b") # singular values and weight vectors
).unnest("a")

df.select(
    pds.query_singular_values("a", "b", center = True, as_explained_var=True)
)

The Xi - Correlation is also implemented:

df.select(
    pds.xi_corr("x", "y")
)

What's Changed

• try new maturin config by @abstractqqq in #128
• added basic pca exprs by @abstractqqq in #129
• added infer_prob by @abstractqqq in #131
• More pca by @abstractqqq in #132
• xi_corr by @abstractqqq in #134

Full Changelog: v0.4.1-release...v0.4.2

v0.4.1-release

Original release note

What's Changed

• upgraded download action in CI by @abstractqqq in #124
• removed unncessary test-deps by @abstractqqq in #126

Full Changelog: v0.4.1-fix-release...v0.4.1-release

v0.4.1-fix-release

release note

v0.4.1

What's Changed

• updated packages, OLS now falls back to LU decomp when X^TX is not strictly positive definite. by @abstractqqq in #122

v0.4.0 was yanked due to incomplete upload

v0.4.0 Breaking Changes

For almost all old functions that were invoked like

pl.col("a").name_space.method_call(, ...)

You can now call them via

import polars_ds as pds

pds.method_call(...)

Linters now should recognize the package and all its methods.

v0.4.0 What's Changed

• added meta and str_stats for Diagnosis by @abstractqqq in #110
• added iqr in Diagnosis by @abstractqqq in #111
• Refactor F stats by @abstractqqq in #112
• Sampling functionalities by @abstractqqq in #114
• Added weighted corr, etc. by @abstractqqq in #116
• Fix regex SyntaxWarning on Python 3.12+ by @jorenham in #118
• Clean up codebase by @abstractqqq in #119
• added heat map for corr in Diagnosis by @abstractqqq in #121

New Contributors

• @jorenham made their first contribution in #118

Full Changelog: v0.3.5...v0.4.0
Full Changelog: v0.4.0...v0.4.1

v0.4.1-test-new-ci

add glob pattern

v0.4.0

Breaking Changes

For almost all old functions that were invoked like

pl.col("a").name_space.method_call(, ...)

You can now call them via

import polars_ds as pds

pds.method_call(...)

Linters now should recognize the package and all its methods.

What's Changed

• added meta and str_stats for Diagnosis by @abstractqqq in #110
• added iqr in Diagnosis by @abstractqqq in #111
• Refactor F stats by @abstractqqq in #112
• Sampling functionalities by @abstractqqq in #114
• Added weighted corr, etc. by @abstractqqq in #116
• Fix regex SyntaxWarning on Python 3.12+ by @jorenham in #118
• Clean up codebase by @abstractqqq in #119
• added heat map for corr in Diagnosis by @abstractqqq in #121

New Contributors

• @jorenham made their first contribution in #118

Full Changelog: v0.3.5...v0.4.0

v0.3.5

Breaking Changes:

1. Now there are two types of random column generation methods: 1. With a column reference. These methods are renamed from sample_xxx to rand_xxx. They behave the same as the old way and must be called with a reference column. New methods generate random columns without any reference, therefore new methods won't respect null and won't use reference's statistics. But new methods are easier to use. Using new methods to get a random df.

import polars as pl
import polars_ds as pds

df = pds.random_data(size=100_000, n_cols = 1).select(
    pds.random(0.0, 12.0).alias("uniform_1"),
    pds.random(0.0, 1.0).alias("uniform_2"),
    pds.random_exp(0.5).alias("exp"),
    pds.random_normal(0.0, 1.0).alias("normal"),
    pds.random_normal(0.0, 1000.0).alias("fat_normal"),
)
df.head()

What's Changed

• Knn entropy by @abstractqqq in #104
• diagonsis basics by @abstractqqq in #105
• Better stats by @abstractqqq in #106
• Add profile by @abstractqqq in #107
• added exclude in dependency plots by @abstractqqq in #108
• add Transfer Entropy and related measures by @remiadon in #84

New Contributors

• @remiadon made their first contribution in #84

Full Changelog: v0.3.4-fix-release...v0.3.5