---
url: >-
  https://jer-k.github.io/effect-prophet/benchmarks/linear/map-training-duplicate-features-auto.md
---

# Repeated timestamps with features

 [← All linear growth benchmarks](/benchmarks/linear)

Repeated dates whose regressor and condition values differ from row to row.

## Input

[`benchmark/inputs/v1/training-duplicate-features.json`](https://github.com/jer-k/effect-prophet/blob/main/benchmark/inputs/v1/training-duplicate-features.json): 96 history rows and 8 forecast rows.

## Output

Largest difference: **0.00093** (0.017% of the input's range).

Every forecast row from both libraries.

| Date | Effect Prophet | Python Prophet | Difference |
| --- | ---: | ---: | ---: |
| 2020-01-25 | 17.0806 | 17.0808 | 0.00019 |
| 2020-01-26 | 16.4980 | 16.4981 | 0.00014 |
| 2020-01-27 | 16.0559 | 16.0562 | 0.00033 |
| 2020-01-28 | 16.4228 | 16.4231 | 0.00031 |
| 2020-01-29 | 17.7144 | 17.7148 | 0.00038 |
| 2020-01-30 | 17.4415 | 17.4421 | 0.00059 |
| 2020-01-31 | 18.3572 | 18.3579 | 0.00069 |
| 2020-02-01 | 18.2191 | 18.2200 | 0.00093 |

## Timing

Median of all measured repetitions.

| Step | Effect Prophet | Python Prophet | Effect Prophet is |
| --- | ---: | ---: | --- |
| Fit | 31.9 ms | 113 ms | 3.6× faster |
| Predict | 241 µs | 3.42 ms | 14× faster |
| Save model to JSON | 20 µs | 469 µs | 24× faster |
| Load model from JSON | 260 µs | 2.60 ms | 10× faster |
| Load and predict in a new process | 163 ms | 1.44 s | 8.9× faster |

## Settings

The case declaration both libraries ran.

```json
{
  "kind": "map",
  "comparison": {
    "kind": "equivalent-objective",
    "evidenceId": "linear-output-first-v4"
  },
  "configuration": {
    "growth": "linear",
    "scaling": "absmax",
    "seasonalityMode": "additive",
    "holidaysMode": "additive",
    "changepoints": {
      "mode": "auto",
      "count": 25,
      "range": 0.8
    },
    "changepointPriorScale": 0.05,
    "seasonalities": [
      {
        "name": "weekly-active",
        "periodDays": 7,
        "fourierOrder": 1,
        "priorScale": 10,
        "conditionName": "active"
      }
    ],
    "events": [],
    "regressors": [
      {
        "name": "promotion",
        "priorScale": 10,
        "standardization": "never"
      }
    ]
  },
  "effectOptimizer": {
    "algorithm": "auto",
    "maxIterations": 10000,
    "fallback": "newton",
    "lbfgs": {
      "historySize": 5,
      "initAlpha": 0.001,
      "tolObj": 1e-12,
      "tolRelObj": 10000,
      "tolGrad": 1e-8,
      "tolRelGrad": 10000000,
      "tolParam": 1e-8
    }
  },
  "pythonOptimizer": {
    "algorithm": "Auto",
    "maxIterations": 10000,
    "newtonFallback": true,
    "sigFigs": 12
  }
}
```