https://github.com/erp12/rica
DataFrame abstraction for Clojure data scientists.
https://github.com/erp12/rica
clojure clojurescript data-science dataframe
Last synced: 28 days ago
JSON representation
DataFrame abstraction for Clojure data scientists.
- Host: GitHub
- URL: https://github.com/erp12/rica
- Owner: erp12
- License: epl-1.0
- Created: 2018-06-18T18:23:30.000Z (almost 7 years ago)
- Default Branch: master
- Last Pushed: 2018-11-28T00:46:22.000Z (over 6 years ago)
- Last Synced: 2025-03-25T10:22:00.794Z (about 2 months ago)
- Topics: clojure, clojurescript, data-science, dataframe
- Language: Clojure
- Homepage: https://erp12.github.io/Rica/
- Size: 104 KB
- Stars: 6
- Watchers: 3
- Forks: 0
- Open Issues: 4
-
Metadata Files:
- Readme: README.md
- Changelog: CHANGELOG.md
- License: LICENSE
Awesome Lists containing this project
README
# Rica
[](https://circleci.com/gh/erp12/Rica)
Data-frame abstraction for Clojure data scientists.
A Data-frame is a data structure where **r**ows are accessed by **i**ndex and **c**olumns are **a**ssociated with names. Hence the name **Rica**.
Rica offers a data-frame abstraction that can be broken down into two parts. First, a DataFrame type that implements a handful of Clojure interfaces and protocols and is compatible with most clojure functions that manipulate data-structures. Second, `rica.core` contains functions that make up a data-frame API inspired by [SparkSQL](https://spark.apache.org/sql/), [Pandas](https://pandas.pydata.org/), and [R](https://www.r-project.org/).
Rica aims to offer an intuavite abstraction for manipulating structured data, without requiring a special runtimes (like Spark). The consequence of this is that Rica is not (currently) ideal for large dataset, however it can be useful for smaller tasks and interactive data exploration.
## Usage
Rica has not been put on Clojars yet.
### Creating DataFrames
The `create-data-frame` function can create data-frames in a way that is similar to maps. Every other argument is a keyword that denotes a column name. The following argument is a collection (list of vector) denoting the values stored in that column.
```clojure
(require '[rica.core :refer :all])
; nil
(def df
(create-data-frame :name ["alice" "bob" "eddie"]
:age [32 nil 24]))
; #'data-frame.core/df
(print-schema df)
; root
; |-- :name : java.lang.String
; |-- :age : java.lang.Long
; nil
(show df)
; | :name | :age |
; |-------+------|
; | alice | 32 |
; | bob | |
; | eddie | 24 |
; nil
```
Often times datasets are stored as nested data structres. Rica provides 2 functions to
### Some Example Pipelines
Suppose we generate the following example dataset.
```clojure
(require '[data-frame.core :refer :all])
(def sales
(create-data-frame
:sale-id [0 1 2 4 5 6 7 8 9]
:amount [74.24 31.24 10.40 22.78 58.24 14.26 61.56 13.95 58.15 69.87]
:num-items [3 2 1 2 3 3 2 1 3 1]
:customer-id ["A" "C" "B" "A" "D" nil "B" "E" "A" "F"]))
```
The below snippet finds the biggest purchase made by customer "A".
```clojure
(-> sales
(where #(= "A" (:customer-id %)))
(order-by :amount)
last
println)
; #ordered/map ([:customer-id A] [:amount 58.15] [:num-items 3] [:sale-id 8])
```
The below snippet finds the average cost of individual items purchased in each sale.
```clojure
(-> sales
(select :amount :num-items)
(with-column :avg-item-cost #(/ (:amount %) (:num-items %)))
show)
; | :amount | :num-items | :avg-item-cost |
; |---------+------------+--------------------|
; | 34.24 | 3 | 11.413333333333334 |
; | 31.24 | 2 | 15.62 |
; | 10.4 | 1 | 10.4 |
; | 22.78 | 2 | 11.39 |
; | 58.24 | 3 | 19.413333333333334 |
; | 14.26 | 3 | 4.753333333333333 |
; | 61.56 | 2 | 30.78 |
; | 13.95 | 1 | 13.95 |
; | 58.15 | 3 | 19.383333333333333 |
; | 69.87 | 1 | 69.87 |
; nil
```
### Full API
Tutorial guides and full documenation are still under constrctions, however
the full API provided by all Rica namespaces can be found on
[Rica's GitHub Pages site](https://erp12.github.io/Rica/index.html).
Generated using [Codox](https://github.com/weavejester/codox) Leiningen plugin.
## Road Map / To Do
For a complete list of features being worked on, see the [project's GitHub issues](https://github.com/erp12/Rica/issues).
If you have any requests or encounter any issues using Rica, please open a GitHub issue (or pull request). A more complete contributing guide is coming soon.
## License
Copyright © 2018 Edward Pantridge
Distributed under the Eclipse Public License either version 1.0 or (at
your option) any later version.