程式碼解釋器範例

In [64]: !head -n1 any20240529.csv
,分組,對話次數,參與人數,文件詞組,port
In [65]: !tail -n1 any20240529.csv
14,合計,750+,59+,"3,357,487+",NaN

from openai import OpenAI
client = OpenAI()

fname='any20240529.csv'
file = client.files.create(
  file=open(fname, "rb"),
  purpose='assistants'
)

assistant = client.beta.assistants.create(
  instructions="You are a personal python tutor. When asked a coding question, write and run code to answer the question.",
  model="gpt-4o",
  tools=[{"type": "code_interpreter"}],
  tool_resources={
    "code_interpreter": {
      "file_ids": [file.id]
    }
  }
)

def run_cmd(cmd):
  thread = client.beta.threads.create(
    messages=[
    {
      "role": "user",
      "content": "I need to solve the problem \`"+cmd+"\`. Can you help me?",
      "attachments": [
        {
          "file_id": file.id,
          "tools": [{"type": "code_interpreter"}]
        }
      ]
    }
  ]
  )

  stream = client.beta.threads.runs.create(
    thread_id=thread.id,assistant_id=assistant.id,stream=True
  )
  for event in stream:
    if  event.event=='thread.message.completed': print(event.data.content[0].text.value)

  return

In [66]: run_cmd("這個數據集的基本資訊?")
In [67]: run_cmd("哪個分組有最高的「參與人數」?")
In [68]: run_cmd("哪各分組有最高的「參與人數」，且分組名稱不等於「合計」?")
In [69]: run_cmd("哪個分組有最高的「參與人數」，注意欄位的值有可能有加號(「+」)的干擾，且分組名稱不等於「合計」?")

In [69]: run_cmd("哪個分組有最高的「參與人數」，注意欄位的值有可能有加號(「+」)的干擾，且分組名稱不等於「合計」?")
Sure, let's start by inspecting the contents of the uploaded file to understand its structure. This will help us to solve the problem by identifying the group with the highest "participant count."

We will load and display the first few rows of the file to get an idea of its format.
It seems the file format is not recognized as CSV or Excel. Let's check the exact name and extension of the file to identify its type.
The file does not have an extension, which is probably why it couldn’t be identified properly. Let’s try reading the file as a CSV with different delimiters and determine its structure.

We will read the first few lines as text to better understand how to parse it.
The file appears to be in CSV format, and it looks like it uses commas as delimiters. Additionally, certain columns have "+" signs and spaces within the values.

Here is the plan:

1. Load the file as a CSV, handling potential spaces and "+" signs in the participant counts.
2. Filter out any rows where the group name is "合計".
3. Convert the "參與人數" column to numeric values by removing any "+" signs.
4. Identify the group with the highest participant count.

Let's perform these steps in code.
The dataframe has been loaded successfully. Now, let's focus on cleaning and processing the "參與人數" column to identify the group with the highest participant count. We will remove any "+" signs and convert the values to numeric. Also, we will filter out rows where the group name is "合計".
The group with the highest number of participants is \( 研發及資訊部 \) with 7 participants.

If you need any further analysis or assistance, feel free to ask!

for event in stream:
    if event.event=='thread.run.step.completed':
        a=event.data.step_details
        if a.type!="message_creation":print(a.tool_calls[0].code_interpreter.input)

In [71]: run_cmd("產生此數據集的各個欄位變數的統計圖形檔案")

import pandas as pd

​# Load the dataset

file_path = '/mnt/data/file-pkHahW6gd2M88UV1C3vf0xjx'
data = pd.read_csv(fname)

# Display the first few rows of the dataset and its info to understand its structure
data.head(), data.info()

import seaborn as sns
import matplotlib.pyplot as plt

# Clean the 'number of conversations', 'number of participants', and 'file phrase' columns
columns_to_clean = [
    ' number of conversations', 
    ' number of participants', 
    ' file phrase'
]

# Remove non-numeric characters and convert to numeric

for col in columns_to_clean:
    data[col] = pd.to_numeric(data[col].str.replace('[^0-9]', '', regex=True))

# Check the cleaned data
data.head()
# Setup the plotting environment and specify the columns for plotting

sns.set(style="whitegrid")

columns_to_plot = [' number of conversations', ' number of participants', ' file phrase', ' port']

​# Generate the plots
fig, axes = plt.subplots(nrows=4, ncols=2, figsize=(15, 20))

for i, col in enumerate(columns_to_plot):
    # Histogram
    sns.histplot(data[col], kde=True, ax=axes[i, 0])
    axes[i, 0].set_title(f'Histogram of {col.strip()}')
    # Boxplot
    sns.boxplot(x=data[col], ax=axes[i, 1])
    axes[i, 1].set_title(f'Boxplot of {col.strip()}')

plt.tight_layout()
plt.show()

In [72]: run_cmd("產生此數據集的各個欄位變數的統計圖形檔案，使用streamlit套件，生成程式碼。")

kuang@eng06 /nas2/kuang/MyPrograms/query_anything
$ cat plot4.py
import pandas as pd
import os

# Load the uploaded file
file_path = 'any20240529.csv'
data = pd.read_csv(file_path)

# Display the first few rows and summary information of the dataset
data_info = data.info()
data_head = data.head()

data_info, data_head
# Define the Streamlit script content
streamlit_script = """
import streamlit as st
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns

plt.rcParams['font.sans-serif'] = ['SimHei']  # 設置中文字體為SimHei
plt.rcParams['axes.unicode_minus'] = False  # 解決負數標籤顯示為方塊的問題

# Load the data
uploaded_file = st.file_uploader("Choose a CSV file", type="csv")
if uploaded_file is not None:
    data = pd.read_csv(uploaded_file)

    # Preprocess the data
    data['對話次數'] = data['對話次數'].apply(lambda x: x.replace('+', '')).astype(float)
    data['參與人數'] = data['參與人數'].apply(lambda x: x.replace('+', '')).astype(float)
    data['文件詞組'] = data['文件詞組'].apply(lambda x: x.replace('+', ''))
    data['文件詞組'] = data['文件詞組'].apply(lambda x: x.replace(',', '')).astype(float)

    # Handling missing values
    data = data.fillna(0)  # Filling missing values with 0 for simplicity

    st.write("Data Preview:", data.head())

    # Generate plots for each column
    st.write("### Statistical Plots")

    for column in data.columns:
        if column not in ['Unnamed: 0', '分組']:
            st.write(f"#### {column}")
            fig, ax = plt.subplots()
            sns.histplot(data[column], kde=True, ax=ax)
            st.pyplot(fig)
"""
script_path = './streamlit_app.py'
with open(script_path, 'w') as file:
    file.write(streamlit_script)

os.system('streamlit run '+ script_path + '&')