Skip to content

Commit e47955c

Browse files
ajayarunachalamajayarunachalam
authored
Add files via upload
0 parents  commit e47955c

38 files changed

+14303
-0
lines changed

‎README.rst

+208
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,208 @@
1+
`GUIPandasAI`: An open-source, low-code simple python wrapper around PandasAI using the Streamlit Framework.
2+
============================================================================================================
3+
4+
`gui-pandas-ai` is concieved, designed and developed by `Ajay Arunachalam` (ajay.arunachalam08@gmail) - https://www.linkedin.com/in/ajay-ph-d-4744581a/
5+
6+
**gui-pandas-ai**
7+
8+
pypi: https://pypi.org/project/gui-pandas-ai
9+
10+
11+
The APP workflow is as seen below.
12+
.. image:: images/main_ui.png
13+
14+
The users after sucessful login, are redirected to the API key input window to submit their respective openAI key. Next, the users can upload their flat csv file followed by their data analysis queries. The history of the prompts and responses can be stored in the text file, along with any response plots. Simply, one can ask questions about your data and get the answers back, in the form of human natural language response.
15+
16+
17+
About `GUIPandasAI`
18+
===================
19+
20+
`gui-pandas-ai` is a simple, ease-to-use Python UI Wrapper built to use `PandasAI` as naively and intuitively as possible. `gui-pandas-ai` provides an easy web gui interface to access `ChatGPT` directly along with provision for several key data analysis utilities. It is altogether a `low-code` solution. With this utility APP one can perform all end-to-end data analysis simply with text-based input queries democratizing Generative AI functionalities. User's can simply ask question related to their data and get the corresponding analysis as response. Further, one can also get quick insights, explore trends & patterns, get the aggregated results, fetch data profiling report and data summary, rendered SQL view of data for offline SQL analysis, data storytelling extract, etc.
21+
22+
GUIPandasAI Usage Steps
23+
=======================
24+
Step 1) Create a virtual environment
25+
26+
.. code:: bash
27+
28+
py -3 -m venv <your_env_name>
29+
cd <your_env_name>/Scripts/activate
30+
31+
**or**
32+
33+
conda create -n <your_env_name> python=3.x (or 3.x)
34+
source activate <your_env_name>
35+
36+
Step 2) Create the clone of the repository in your created virtual environment
37+
38+
.. code:: bash
39+
40+
>>> git clone https://github.com/ajayarunachalam/gui-pandas-ai
41+
>>> cd gui-pandas-ai
42+
>>> pip install -r requirements.txt
43+
44+
$ git clone https://github.com/ajayarunachalam/gui-pandas-ai
45+
$ cd gui-pandas-ai
46+
  $ sudo bash setup.sh
47+
48+
**or**
49+
$ git clone https://github.com/ajayarunachalam/gui-pandas-ai
50+
  $ cd gui-pandas-ai
51+
  $ sudo bash setup.sh or python setup.py install
52+
53+
Step 3) Launch APP
54+
55+
- **The users can set their own credentials in the file `secrets.toml` found under the folder `.streamlit`. Alternatively, one can also use the existing example credentials as-it-is which are `user_test` and `user@123` or
56+
`dev_test` and `dev@123`**
57+
58+
- **Windows users within the cloned folder just simply double-click the "run_app_windows.bat" file. Note:- Open the file with an Editor and replace with your virtual directory path within the file**
59+
60+
- **Linux users navigate within the cloned folder and type in "sudo bash run_app_linux.sh" in the terminal**
61+
62+
- **Mac users navigate within the cloned folder and type in "sh run_app_mac.sh" in the terminal**
63+
64+
The APP will launch with a URL as seen below.
65+
66+
.. image:: images/run_app.png
67+
68+
69+
APP Q&A Window
70+
==============
71+
As seen below the user's can drag and drop their `CSV` files or upload them, and submit their questions in form of simple queries. The data analysis results are received back in the form of natural language.
72+
73+
.. image:: images/page0.png
74+
75+
GUIPandasAI Code Snippet
76+
========================
77+
Below is the example code snippet that runs the LLMs while viewing the uploaded data.
78+
79+
```python
80+
if st.session_state.df is not None:
81+
st.subheader("Peek into the uploaded dataframe:")
82+
st.write(st.session_state.df.head(2))
83+
84+
with st.form("Question"):
85+
question = st.text_area("Question", value="", help="Enter your queries here")
86+
answer = st.text_area("Answer", value="")
87+
submitted = st.form_submit_button("Submit")
88+
if submitted:
89+
with st.spinner():
90+
llm = OpenAI(api_token=st.session_state.openai_key)
91+
pandas_ai = PandasAI(llm)
92+
x = pandas_ai.run(st.session_state.df, prompt=question)
93+
94+
fig = plt.gcf()
95+
fig, ax = plt.subplots(figsize=(10, 6))
96+
plt.tight_layout()
97+
if fig.get_axes() and fig is not None:
98+
st.pyplot(fig)
99+
fig.savefig("plot.png")
100+
st.write(x)
101+
st.session_state.prompt_history.append(question)
102+
response_history.append(x) # Append the response to the list
103+
st.session_state.response_history = response_history
104+
```
105+
106+
PandasAI - Overview
107+
===================
108+
`Pandas AI` is a Python library that adds generative artificial intelligence capabilities to Pandas, the popular data analysis and manipulation tool. `PandasAI` [PandasAI](https://github.com/gventuri/pandas-ai) aims to make Pandas dataframes conversational, allowing you to ask questions about your data and get answers back, in the form of natural human language.
109+
110+
For quick overview glimse through the below illustration: (All Credits & Copyrights Reserved to `Pandas AI`)
111+
112+
```python
113+
import pandas as pd
114+
from pandasai import PandasAI
115+
116+
# Sample DataFrame
117+
df = pd.DataFrame({
118+
"country": ["United States", "United Kingdom", "France", "Germany", "Italy", "Spain", "Canada", "Australia", "Japan", "China"],
119+
"gdp": [19294482071552, 2891615567872, 2411255037952, 3435817336832, 1745433788416, 1181205135360, 1607402389504, 1490967855104, 4380756541440, 14631844184064],
120+
"happiness_index": [6.94, 7.16, 6.66, 7.07, 6.38, 6.4, 7.23, 7.22, 5.87, 5.12]
121+
})
122+
123+
# Instantiate a LLM
124+
from pandasai.llm.openai import OpenAI
125+
llm = OpenAI(api_token="YOUR_API_TOKEN")
126+
127+
pandas_ai = PandasAI(llm, conversational=True)
128+
pandas_ai(df, prompt='Which are the 5 happiest countries?')
129+
```
130+
131+
The above code will return the following:
132+
133+
```
134+
6 Canada
135+
7 Australia
136+
1 United Kingdom
137+
3 Germany
138+
0 United States
139+
Name: country, dtype: object
140+
```
141+
142+
Of course, you can also ask PandasAI to perform more complex queries. For example, you can ask PandasAI to find the sum of the GDPs of the 2 unhappiest countries:
143+
144+
```python
145+
pandas_ai(df, prompt='What is the sum of the GDPs of the 2 unhappiest countries?')
146+
```
147+
148+
The above code will return the following:
149+
150+
```
151+
19012600725504
152+
```
153+
154+
```python
155+
"""Example of using PandasAI on multiple Pandas DataFrame"""
156+
157+
import pandas as pd
158+
from pandasai import PandasAI
159+
from pandasai.llm.openai import OpenAI
160+
161+
employees_data = {
162+
'EmployeeID': [1, 2, 3, 4, 5],
163+
'Name': ['John', 'Emma', 'Liam', 'Olivia', 'William'],
164+
'Department': ['HR', 'Sales', 'IT', 'Marketing', 'Finance']
165+
}
166+
167+
salaries_data = {
168+
'EmployeeID': [1, 2, 3, 4, 5],
169+
'Salary': [5000, 6000, 4500, 7000, 5500]
170+
}
171+
172+
employees_df = pd.DataFrame(employees_data)
173+
salaries_df = pd.DataFrame(salaries_data)
174+
175+
176+
llm = OpenAI()
177+
pandas_ai = PandasAI(llm, verbose=True)
178+
response = pandas_ai([employees_df, salaries_df], "Who gets paid the most?")
179+
print(response)
180+
```
181+
182+
```
183+
# Output: Olivia
184+
```
185+
186+
License
187+
=======
188+
Copyright 2022-2023 Ajay Arunachalam <[email protected]>
189+
190+
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:
191+
192+
The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.
193+
194+
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. © 2023 GitHub, Inc.
195+
196+
References
197+
==========
198+
Special mention to `streamlit`, `openai`, `PandasAI`, `Pandas Profiling` and the other open-source communities for their incredible contributions.
199+
200+
201+
TODO
202+
====
203+
204+
- [ ] Include more LLMs
205+
- [ ] Add support for Big Data
206+
- [ ] Add Statistical data analysis
207+
- [ ] Add Adv. Data Analytics provision
208+
- [ ] Integrate Lux based visualizations

‎blog_examples/TODO.txt

Whitespace-only changes.

‎data/gdp.csv

+11
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,11 @@
1+
"country","gdp","happiness_index"
2+
"United States",21400000,7.3
3+
"United Kingdom",2940000,7.2
4+
"France",2830000,6.5
5+
"Germany",3870000,7
6+
"Italy",2160000,6
7+
"Spain",1350000,6.3
8+
"Canada",1780000,7.3
9+
"Australia",1320000,7.3
10+
"Japan",516000,5.9
11+
"China",14000000,5

0 commit comments

Comments
 (0)