Merge pull request #31 from abstractqqq/docs

Docs

Cargo.toml

@@ -17,8 +17,6 @@ num = "0.4.1"
 faer = {version = "0.15", features = ["ndarray", "nightly"]}
 ndarray = "0.15.6" # see if we can get rid of this
 hashbrown = {version = "0.14.2", features=["nightly"]}
-# Try realfft instead, which is a wrapper around rustfft but specializes for reals and seems to have better perf
-# rustfft = "6.1.0" 
 itertools = "0.12.0"
 aho-corasick = "1.1"
 rand = {version = "0.8.5"} # Simd support feature seems to be broken atm

docs/docs/complex_ext.md

@@ -0,0 +1,3 @@
+## Extension for Complex Numbers
+
+::: polars_ds.complex_ext

docs/docs/index.md

@@ -0,0 +1,44 @@
+# Polars-ds
+
+A Polars Plugin aiming to simplify common numerical/string data analysis procedures. This means that the most basic data science, stats, NLP related tasks can be done natively inside a dataframe, without leaving dataframe world. This also means that for simple data pipelines, you do not need to install NumPy/Scipy/Scikit-learn, which saves a lot of space, which is great under constrained resources.
+
+Its goal is NOT to replace SciPy, or NumPy, but rather it tries reduce dependency for simple analysis, and tries to reduce Python side code and UDFs, which are often performance bottlenecks.
+
+## Getting Started
+```bash
+pip install polars_ds
+```
+
+and 
+
+```python
+import polars_ds
+```
+when you are using the namespaces provided by the package.
+
+## Examples
+
+Generating random numbers, and running t-test, normality test inside a dataframe
+```python
+df.with_columns(
+    pl.col("a").stats_ext.sample_normal(mean = 0.5, std = 1.).alias("test1")
+    , pl.col("a").stats_ext.sample_normal(mean = 0.5, std = 2.).alias("test2")
+).select(
+    pl.col("test1").stats_ext.ttest_ind(pl.col("test2"), equal_var = False).alias("t-test")
+    , pl.col("test1").stats_ext.normal_test().alias("normality_test")
+).select(
+    pl.col("t-test").struct.field("statistic").alias("t-tests: statistics")
+    , pl.col("t-test").struct.field("pvalue").alias("t-tests: pvalue")
+    , pl.col("normality_test").struct.field("statistic").alias("normality_test: statistics")
+    , pl.col("normality_test").struct.field("pvalue").alias("normality_test: pvalue")
+)
+```
+
+Blazingly fast string similarity comparisons. (Thanks to [RapidFuzz](https://docs.rs/rapidfuzz/latest/rapidfuzz/))
+```python
+df2.select(
+    pl.col("word").str_ext.levenshtein("world", return_sim = True)
+).head()
+```
+
+And a lot more!

docs/docs/num_ext.md

@@ -0,0 +1,3 @@
+## Extension for General Numerical Features/Metrics/Quantities
+
+::: polars_ds.num_ext

docs/docs/stats_ext.md

@@ -0,0 +1,3 @@
+## Extension for Statistical Tests and Samples
+
+::: polars_ds.stats_ext

docs/docs/str_ext.md

@@ -0,0 +1,3 @@
+## Extension for String Manipulation and Metrics
+
+::: polars_ds.str_ext

docs/mkdocs.yml

@@ -0,0 +1,20 @@
+site_name: Polars-ds Docs
+
+nav:
+- Home: index.md
+- Complex Extension: complex_ext.md
+- Numerical Extension: num_ext.md
+- Stats Extension: stats_ext.md
+- String Extension: str_ext.md
+
+theme:
+  name: material
+
+plugins:
+- search
+- mkdocstrings:
+    handlers:
+      python:
+        paths: [../python]
+        selection:
+          docstring_style: numpy

docs/requirements-docs.txt

@@ -0,0 +1,4 @@
+mkdocs
+mkdocstrings[python]
+mkdocs-material
+pytkdocs[numpy-style]