TypeError: float() argument must be a string or a number, not 'Timestamp'

[Youbadawy]

Please make sure that this is a Bug or a Feature Request and provide all applicable information asked by the template. If your issue is an implementation question, please ask your question on StackOverflow or on the TensorTrade Discord #help-desk channel instead of opening a GitHub issue.

System information

• Have I written custom code (as opposed to using example directory):
• OS Platform and Distribution (e.g. Mac Sierra, Windows 10, Linux Ubuntu 16.04):
• Ubuntu 18.04.5 LTS
• TensorTrade version:
• Currently selected TF version: 2.x
• TensorFlow version:
• Python version:
• 3.7.10

You can obtain the TensorFlow version with:
python -c "import tensorflow as tf; print(tf.GIT_VERSION, tf.VERSION)"

Describe the current behavior After initializing all Datastreams, portfolios, and render feeds, and inputting it inside a DQN agent to train,

it receives an error of Timestamp is not float, although we I used the below format to input my data streams as indicated in notebook examples,

for i, count in enumerate(data):
  pair = count.replace("/USD","")
  with NameSpace("bitfinex"):
    if i==0:
      bitfinex_streams = [
          Stream.source(list(data.get(count)[c]), dtype="float").rename(f"{pair}:{c}") for c in data.get(count).columns
      ]
    else:
      bitfinex_streams += [
          Stream.source(list(data.get(count)[c]), dtype="float").rename(f"{pair}:{c}") for c in data.get(count).columns
      ]

feed = DataFeed(bitfinex_streams)

Describe the expected behavior

It should be able to train as the notebook setup_environment_tutorial.ipynb displays a similar format,

Code to reproduce the issue
Provide a reproducible test case that is the bare minimum necessary to generate the problem.

cdd = CryptoDataDownload()
data = {}

data["BTC/USD"] = cdd.fetch("Bitfinex", "USD", "BTC", "1h")
data["ETH/USD"] = cdd.fetch("Bitfinex", "USD", "ETH", "1h")
data["LTC/USD"] = cdd.fetch("Bitfinex", "USD", "LTC", "1h")

data

for i, count in enumerate(data):
  pair = count.replace("/USD","")
  with NameSpace("bitfinex"):
    if i==0:
      bitfinex_streams = [
          Stream.source(list(data.get(count)[c]), dtype="float").rename(f"{pair}:{c}") for c in data.get(count).columns
      ]
    else:
      bitfinex_streams += [
          Stream.source(list(data.get(count)[c]), dtype="float").rename(f"{pair}:{c}") for c in data.get(count).columns
      ]

feed = DataFeed(bitfinex_streams)

bitfinex_options = ExchangeOptions(commission=0.0075, min_trade_price=10.0)
bitfinex = Exchange("bitfinex", service=execute_order,  options=bitfinex_options)(
    Stream.source(list(data.get("BTC/USD")["close"]), dtype="float").rename("USD-BTC"),
    Stream.source(list(data.get("ETH/USD")["close"]), dtype="float").rename("USD-ETH"),
    Stream.source(list(data.get("LTC/USD")["close"]), dtype="float").rename("USD-LTC")
)

portfolio = Portfolio(USD, [
    Wallet(bitfinex, 100000 * USD),
    Wallet(bitfinex, 0 * BTC),
    Wallet(bitfinex, 0 * ETH),
    Wallet(bitfinex, 0 * LTC)
])

env = default.create(
    portfolio=portfolio,
    action_scheme='managed-risk',  # ['managed-risk', 'simple', BSH ]
    reward_scheme='risk-adjusted', # [ 'risk-adjusted', 'simple' ]
    feed=feed,
    renderer_feed=renderer_feed,
    renderer=default.renderers.PlotlyTradingChart(),
    window_size=5
)

agent = DQNAgent(env)

agent.train(n_steps=200, n_episodes=5, save_path="agents/", render_interval=None)

Other info / logs
Include any logs or source code that would be helpful to diagnose the problem. If including tracebacks, please include the full traceback. Large logs and files should be attached.

[daecepi]

@Youbadawy i am having the same error, did you resolve it?

[Youbadawy]

@daecepi yes, use unix timestamp instead, it should work!

[abstractguy]

Please start from the latest examples. This is fixed there. I may merge the fix to data/cdd.py later.