Merge branch 'master' of https://github.com/thomasm1/python_2018

Dockerfile

@@ -0,0 +1,8 @@
+FROM python:3.9-slim
+
+WORKDIR /app
+
+COPY simple_server.py .
+
+CMD ["python", "simple_server.py"]
+

PY_AGENTS/.gitignore

@@ -0,0 +1,2 @@
+node_modules
+.env

PY_AGENTS/agent-mapl-anthropic/.dockerignore

@@ -0,0 +1,4 @@
+.env
+.gitignore
+env
+

PY_AGENTS/agent-mapl-anthropic/Dockerfile

@@ -0,0 +1,18 @@
+FROM python:3.11-slim
+
+WORKDIR /app
+
+COPY . /app
+
+RUN pip install --no-cache-dir -r requirements.txt
+
+#COPY agents.py prompts.py ./
+
+RUN mkdir /app/data
+
+
+EXPOSE 80
+#set the environmen variable 
+#ENV PYTHONUNBUFFERED=1 
+
+CMD ["python", "agents.py"]

PY_AGENTS/agent-mapl-anthropic/README.md

@@ -0,0 +1,2 @@
+docker run -it -v "C:\Users\thoma\git\_python\PY_AGENTS:/app/data" mapl-agents
+

PY_AGENTS/agent-mapl-anthropic/agents.py

@@ -0,0 +1,117 @@
+# agent-mapl.py
+
+import os
+import csv
+import anthropic
+from prompts import *
+
+#KEY
+if not os.getenv("ANTHROPIC_API_KEY"):
+  os.environ["ANTHROPIC_API_KEY"] =  input("Please enter your Antrhopic API key: ") # prompt the user for anthro api key
+
+#CLIENT
+client = anthropic.Anthropic()
+sonnet = "claude-3-5-sonnet-20240620"   # model name get from docs.anthropic.com/en/docs/about-claude/models
+
+# Function to read the CSV file form the user
+def read_csv(file_path):
+  data = []
+  with open(file_path, "r", newline="") as csvfile: #open csv in readmode
+    csv_reader = csv.reader(csvfile)  # create csv reader obj
+    for row in csv_reader:
+      data.append(row)  # add each row to the data list
+  return data
+
+# Saving  
+def save_to_csv(data, output_file, headers=None):
+  mode = 'w' if headers else 'a' # set file mode: w write if headers are provide else a append
+  with open(output_file, mode, newline="") as f:
+    writer=csv.writer(f) # create a csv writer object
+    if headers:
+      writer.writerow(headers) # write the headers if provided
+    for row in csv.reader(data.splitlines()): #sp;lie data stirng into rows
+      writer.writerow(row)
+
+
+#my personal librarian agent
+def analyzer_agent(sample_data):
+  message = client.messages.create(
+    model=sonnet,
+    max_tokens=400, #limit response to 400 tokens
+    temperature=0.1, # set low temp for more focused, deterministic output
+    system=ANALYZER_SYSTEM_PROMPT,  # use the predefined system prompt for analyzer
+    messages=[
+      {
+        "role":"user",
+        "content": ANALYZER_USER_PROMPT.format(sample_data=sample_data)
+        # format user prompt with provided smpale data
+      }
+    ]
+  )
+  return message.content[0].text # return text content of tfirst message
+
+# generator agent
+def generator_agent(analysis_result, sample_data, num_rows=30):
+  message = client.messages.create(
+    model=sonnet,
+    max_tokens=1500, # allow for longer response
+    temperature=1, # high temp for creative, divers output
+    system=GENERATOR_SYSTEM_PROMPT,
+    messages=[
+      {
+        "role":"user",
+        "content": GENERATOR_USER_PROMPT.format(
+          num_rows=num_rows,
+          analysis_result=analysis_result,
+          sample_data=sample_data
+        )
+      }
+    ]
+  )    
+  return message.content[0].text
+
+
+#main fucntion to run analyzer and generator agents: CONTROL THE AGENTS!
+
+#get input
+file_path = input("\nPuh-lease enter the name of your CSV file: ")
+file_path = os.path.join('/app/data', file_path)
+desired_rows =  int(input("n now Enter the number of rows you want in the new dataset: "))
+
+#read data
+sample_data = read_csv(file_path)
+sample_data_str  = "\n".join([",".join(row) for row in sample_data]) # convert 2d list to a single string
+
+print("\nLaunching team of Agents")
+
+analysis_result = analyzer_agent(sample_data_str)
+print("\n### Analyzer Agent output:###\n")
+print(analysis_result)
+print("\n=-============\n\nGenerating new data....")
+
+#output
+output_file = "/app/data/new_dataset.csv"
+headers = sample_data[0]
+
+#creat output with headers
+save_to_csv("", output_file, headers)
+
+
+batch_size = 30  #number of rows to genrate in each batch
+generated_rows = 0 # counter to keep track of rows generated
+
+# generate data in batches until rows reached
+while generated_rows < desired_rows:
+  rows_to_generate = min(batch_size, desired_rows - generated rows)
+  generated_data = generator_agent(analysis_result, sample_data_str, rows_to_generate)
+  #append gen data to output file
+  save_to_csv(generated_data, output_file)
+  #update count of generated rows
+  generated_rows += rows_to_generate
+  #print progress update
+  print(f"Generated {generated_rows} rows out of {desired_rows}")
+
+# inform process complete
+print(f"\nGenerated data has been saved to {output_file}")
+
+# THE END

PY_AGENTS/agent-mapl-anthropic/input.csv

@@ -0,0 +1,4 @@
+Type,Indicators,Name,Instructions
+bookwork,studious,iWorm,read and teach
+taskrunner,heavy duty,ichabod,pick up and carry
+

PY_AGENTS/agent-mapl-anthropic/prompts.py

@@ -0,0 +1,32 @@
+#prompts.py
+ANALYZER_SYSTEM_PROMPT = """You are an AI agent that analyzes the CSV provided by the user. 
+The focus of your analysis should be on what the data is, how it is formatted, what each column stands for, and how new data should be interpreted."""
+
+GENERATOR_SYSTEM_PROMPT = """You are an AI agent that generates new CSV rows based on analyssi results and sample data. 
+Follow the exact formatting and don't outpu any extra text. You only output formatted data, never any other text."""
+
+ANALZER_USER_PROMPT = """ Analyze the structure and patterns of this sample dataset:
+
+{sampl_data}
+
+Provide a concise summary of the following: 
+1. formatting of the dataset, be crystal clear when describing the structure of the CSV
+2. what the dataset represets, what each column stands for
+3. how new data should look like, based on the patterns you've identified
+"""
+
+GENERATOR_USER_PROMPT = """Generate {num_rows} new CSV rows based on this analysis and sample data:
+
+
+Analysis:
+{analysis_result}
+
+Sample Data: 
+{sample_data}
+
+Use the exact same formatting as the orginal data. Output onlyh the generated code, no extra text.
+
+DO NOT USE ANY TEXT BEFVOFE/AFTER THE DATA. JUST START BY OUTPUTTING THE NEW ROWS. NO EXTRA TEXT!!!!"""
+
+
+

PY_AGENTS/agent-mapl-anthropic/requirements.txt

@@ -0,0 +1,2 @@
+anthropic
+

PY_AI/blockchain_data.csv

@@ -0,0 +1,11 @@
+Transaction_ID,Timestamp,Sender,Receiver,Amount,Transaction_Fee
+tx_1,2025-01-22 10:00:00,wallet_1,wallet_11,100,0.01
+tx_2,2025-01-22 10:05:00,wallet_2,wallet_12,250,0.015
+tx_3,2025-01-22 10:10:00,wallet_3,wallet_13,300,0.02
+tx_4,2025-01-22 10:15:00,wallet_4,wallet_14,150,0.01
+tx_5,2025-01-22 10:20:00,wallet_5,wallet_15,200,0.025
+tx_6,2025-01-22 10:25:00,wallet_6,wallet_16,350,0.02
+tx_7,2025-01-22 10:30:00,wallet_7,wallet_17,400,0.03
+tx_8,2025-01-22 10:35:00,wallet_8,wallet_18,50,0.01
+tx_9,2025-01-22 10:40:00,wallet_9,wallet_19,75,0.015
+tx_10,2025-01-22 10:45:00,wallet_10,wallet_20,125,0.02

PY_AI/py_torch/blockchain-1.py

@@ -0,0 +1,21 @@
+# dataset.py
+import pandas as pd
+from torch.utils.data import Dataset
+
+class BlockchainDataset(Dataset):
+    def __init__(self, csv_file): 
+        self.data = pd.read_csv(csv_file)
+
+    def __len__(self):
+        return len(self.data)
+
+    def __getitem__(self, idx):
+        row = self.data.iloc[idx]
+        # Extract input features and labels (e.g., predict "Amount")
+        features = {
+            'sender': row['Sender'],
+            'receiver': row['Receiver'],
+            'fee': row['Transaction_Fee'],
+        }
+        label = row['Amount']
+        return features, label

PY_AI/py_torch/blockchain-2.py

@@ -0,0 +1,17 @@
+# pipeline.py
+import torch
+from torch.utils.data import DataLoader
+from sklearn.preprocessing import LabelEncoder
+
+def preprocess_data(dataset):
+    # Encode sender and receiver addresses as integers
+    sender_encoder = LabelEncoder()
+    receiver_encoder = LabelEncoder()
+
+    dataset.data['Sender'] = sender_encoder.fit_transform(dataset.data['Sender'])
+    dataset.data['Receiver'] = receiver_encoder.fit_transform(dataset.data['Receiver'])
+
+    return dataset
+
+def create_dataloader(dataset, batch_size=2):
+    return DataLoader(dataset, batch_size=batch_size, shuffle=True)

PY_AI/py_torch/blockchain-3.py

@@ -0,0 +1,14 @@
+# model.py
+import torch.nn as nn
+
+class TransactionPredictor(nn.Module):
+    def __init__(self, input_dim):
+        super(TransactionPredictor, self).__init__()
+        self.fc = nn.Sequential(
+            nn.Linear(input_dim, 32),
+            nn.ReLU(),
+            nn.Linear(32, 1)  # Output a single value (Amount)
+        )
+
+    def forward(self, x):
+        return self.fc(x)

PY_AI/py_torch/blockchain-4.py

@@ -0,0 +1,36 @@
+# training.py
+import torch
+import torch.optim as optim
+from torch.nn import MSELoss
+from dataset import BlockchainDataset
+from pipeline import preprocess_data, create_dataloader
+from model import TransactionPredictor
+
+csv_file = "blockchain_data.csv"
+dataset = BlockchainDataset(csv_file)
+dataset = preprocess_data(dataset)
+dataloader = create_dataloader(dataset)
+
+# Initialize model, optimizer, and loss function
+model = TransactionPredictor(input_dim=3)  # sender, receiver, fee
+optimizer = optim.Adam(model.parameters(), lr=0.001)
+criterion = MSELoss()
+
+# Training loop
+for epoch in range(10):
+    for batch in dataloader:
+        features, labels = batch
+        sender = features['sender'].float()
+        receiver = features['receiver'].float()
+        fee = features['fee'].float()
+
+        inputs = torch.stack([sender, receiver, fee], dim=1)
+        labels = labels.float()
+
+        optimizer.zero_grad()
+        outputs = model(inputs)
+        loss = criterion(outputs.squeeze(), labels)
+        loss.backward()
+        optimizer.step()
+
+    print(f"Epoch {epoch + 1}, Loss: {loss.item():.4f}")

PY_AI/py_torch/blockchain-5.py

@@ -0,0 +1,19 @@
+# sankey.py
+import matplotlib.pyplot as plt
+import pandas as pd
+from matplotlib.sankey import Sankey
+
+def create_sankey(csv_file):
+    data = pd.read_csv(csv_file)
+
+    sankey = Sankey()
+    for _, row in data.iterrows():
+        sankey.add(flows=[-row['Amount'], row['Amount']],
+                   labels=[row['Sender'], row['Receiver']],
+                   orientations=[-1, 1])
+    sankey.finish()
+    plt.title('Blockchain Transactions Sankey')
+    plt.show()
+
+if __name__ == "__main__":
+    create_sankey("blockchain_data.csv")

PY_AI/py_torch/blockchain_data.csv

@@ -0,0 +1,11 @@
+Transaction_ID,Timestamp,Sender,Receiver,Amount,Transaction_Fee
+tx_1,2025-01-22 10:00:00,wallet_1,wallet_11,100,0.01
+tx_2,2025-01-22 10:05:00,wallet_2,wallet_12,250,0.015
+tx_3,2025-01-22 10:10:00,wallet_3,wallet_13,300,0.02
+tx_4,2025-01-22 10:15:00,wallet_4,wallet_14,150,0.01
+tx_5,2025-01-22 10:20:00,wallet_5,wallet_15,200,0.025
+tx_6,2025-01-22 10:25:00,wallet_6,wallet_16,350,0.02
+tx_7,2025-01-22 10:30:00,wallet_7,wallet_17,400,0.03
+tx_8,2025-01-22 10:35:00,wallet_8,wallet_18,50,0.01
+tx_9,2025-01-22 10:40:00,wallet_9,wallet_19,75,0.015
+tx_10,2025-01-22 10:45:00,wallet_10,wallet_20,125,0.02

PY_AI/py_torch/data_generator.py

@@ -0,0 +1,24 @@
+import pandas as pd
+
+# Mock data for blockchain transactions
+data = {
+    "Transaction_ID": [f"tx_{i}" for i in range(1, 11)],
+    "Timestamp": [
+        "2025-01-22 10:00:00", "2025-01-22 10:05:00", "2025-01-22 10:10:00",
+        "2025-01-22 10:15:00", "2025-01-22 10:20:00", "2025-01-22 10:25:00",
+        "2025-01-22 10:30:00", "2025-01-22 10:35:00", "2025-01-22 10:40:00",
+        "2025-01-22 10:45:00"
+    ],
+    "Sender": [f"wallet_{i}" for i in range(1, 11)],
+    "Receiver": [f"wallet_{i+10}" for i in range(1, 11)],
+    "Amount": [100, 250, 300, 150, 200, 350, 400, 50, 75, 125],
+    "Transaction_Fee": [0.01, 0.015, 0.02, 0.01, 0.025, 0.02, 0.03, 0.01, 0.015, 0.02]
+}
+
+# Create a DataFrame
+blockchain_df = pd.DataFrame(data)
+
+# Save to a CSV file
+file_path = "/mnt/data/blockchain_data.csv"
+blockchain_df.to_csv(file_path, index=False)
+file_path