Accessing underlying backtest data via API

Here's a copy-paste ready script for Kubid, showing how to access portfolio equity (value) at custom points in time using the QuantConnect API, rather than through log files:

# ============================================================
# Quick-start script: download a backtest’s equity curve,
# load it into a pandas DataFrame, then query the portfolio
# value (“Strategy Equity”) at any custom date/time.
#
# 1) Fill-in your API token, project_id, and backtest_id.
# 2) `pip install pandas requests` if they are not available.
# 3) Run – the DataFrame `df` holds the full equity curve.
# ============================================================

from AlgorithmImports import *
import requests                  # HTTP requests to the QC API
import pandas as pd              # Data manipulation
from datetime import datetime

# ------------------------------------------------------------
# REQUIRED: replace the placeholders below
# ------------------------------------------------------------
API_TOKEN   = "<YOUR_API_TOKEN>"     # e.g. generated at https://www.quantconnect.com/account
PROJECT_ID  = "<PROJECT_ID>"         # integer or guid
BACKTEST_ID = "<BACKTEST_ID>"        # guid

# ------------------------------------------------------------
# 1) Build the “read backtest” endpoint
# ------------------------------------------------------------
url = (
    "https://www.quantconnect.com/api/v2/backtests/read"
    f"?project_id={PROJECT_ID}&backtest_id={BACKTEST_ID}"
)

# ------------------------------------------------------------
# 2) Authorize the request
# ------------------------------------------------------------
headers = {
    "Authorization": f"Bearer {API_TOKEN}",
    "Content-Type": "application/json"
}

# ------------------------------------------------------------
# 3) Call the API and parse the JSON response
# ------------------------------------------------------------
response = requests.get(url, headers=headers, timeout=30)
response.raise_for_status()            # raise if the call failed
data = response.json()                 # python dict

# ------------------------------------------------------------
# 4) Pull out the equity-curve points
# ------------------------------------------------------------
equity_values = (
    data["backtest"]                   # root
        ["Charts"]["Strategy Equity"]  # chart
        ["Series"]["Equity"]["Values"] # time-series points
)

# Each point is { "x": 1612137600000, "y": 100000 }  (x = ms epoch)
records: list[tuple[datetime, float]] = []
for pt in equity_values:
    timestamp = datetime.utcfromtimestamp(pt["x"] / 1_000)   # ms → s
    records.append((timestamp, float(pt["y"])))

# ------------------------------------------------------------
# 5) Equity curve → DataFrame with DateTime index
# ------------------------------------------------------------
df = pd.DataFrame(records, columns=["Time", "Equity"]).set_index("Time")

# ------------------------------------------------------------
# 6) Example look-ups
# ------------------------------------------------------------
# a) Exact lookup  (will raise KeyError if exact stamp not present)
try:
    print("Exact 2021-01-04 15:30:00:", df.loc["2021-01-04 15:30:00"])
except KeyError:
    print("Exact timestamp not in DataFrame")

# b) Nearest value *at-or-before* arbitrary time using `asof`
query_time = datetime(2021, 3, 17, 13, 45)
idx = df.index.asof(query_time)        # returns NaT if nothing before this time
portfolio_value = df.loc[idx, "Equity"] if idx is not pd.NaT else None
print(f"Closest ≤ {query_time}: {portfolio_value}")

# c) Slice a date range
march_curve = df["2021-03-01":"2021-03-31"]
print("March 2021 equity curve head:")
print(march_curve.head())

Key steps:

• Fetch the equity curve using the QuantConnect API.
• Convert it into a time-indexed pandas DataFrame.
• Extract equity (portfolio value) at any exact or nearest timestamp using .loc or .asof.

Replace the token and IDs, and Kubid has full, programmatic access to portfolio value at any custom point in time! No logs required.

Sources: Backtest Analysis - QuantConnect.com, US Equity - QuantConnect.com, Live Analysis - QuantConnect.com, Detecting Impactful News In ETF Constituents, Intraday Arbitrage Between Index ETFs

reponse form backtest/read doesnt' have the actual Chart Series Model hence:

equity_values = (
    data["backtest"]                   # root
        ["charts"]["Strategy Equity"]  # chart
        ["Series"]["Equity"]["Values"] # time-series points
)

---------------------------------------------------------------------------
KeyError                                  Traceback (most recent call last)
Cell In[28], line 10
      6 response.raise_for_status()            # raise if the call failed
      7 data = response.json()      
      9 equity_values = (
---> 10     data["backtest"]                   # root
     11         ["charts"]["Strategy Equity"]  # chart
     12         ["Series"]["Equity"]["Values"] # time-series points
     13 )

KeyError: 'Series'

But i managed to pull this with backtest/chart/read

response = requests.post(f'{qc_common.QC_BASE_URL}/backtests/chart/read', headers = qc_common.get_headers(), 
                         json = {"projectId": project_id, "backtestId": backtest_id, "name": "Strategy Equity", "count": 2500})#, timeout=30)
response.raise_for_status()            # raise if the call failed
data = response.json()      
data["chart"]["series"]["Equity"]["values"][0:10]

[[1453266000, 1000000.0, 1000000.0, 1000000.0, 1000000.0],
 [1453967506, 1000000.0, 1000000.0, 1000000.0, 1000000.0],
 [1454669012, 1000000.0, 1006245.15, 937701.11, 945798.7841],
 [1455370518, 940685.505, 947200.645, 900359.035, 939755.6834],
 [1456072025, 939674.535, 997048.375, 939674.535, 978610.4766],
 [1456773531, 990898.945, 1012763.585, 963860.975, 991918.4902],
 [1457475037, 991972.655, 1027444.98, 991972.655, 1007083.3851],
 [1458176544, 1007640.14, 1040285.735, 998922.22, 1037240.802],
 [1458878050, 1040049.275, 1044780.595, 1032696.365, 1038933.7097],
 [1459579556, 1038933.145, 1052208.48, 1035758.815, 1050846.2383]]

But it seems the data returned is weekly - i presume first column is unix epoch?

Questions:
1. Is there anyways i can specify daily?
2. what are the 4 columns after unix epoch? I couldn't find any documentation.

API Error Message: Error retrieving backtest chart, please try again later