-
Notifications
You must be signed in to change notification settings - Fork 394
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
- Loading branch information
Francesco Calcavecchia
committed
Oct 13, 2023
1 parent
bdafabb
commit e89fabf
Showing
4 changed files
with
197 additions
and
0 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,193 @@ | ||
# Evidently | ||
|
||
DVCLive can be used to track the results | ||
of[Evidently](https://www.evidentlyai.com/). In the following we demonstrate it | ||
through an example. | ||
|
||
[π Google Colab notebook](https://colab.research.google.com/drive/14usegPOSArF9tdO7NUOPndNk5vuUaYaX?usp=sharing) | ||
|
||
## Setup | ||
|
||
```cli | ||
$ pip install dvc dvclive evidently pandas | ||
``` | ||
|
||
## Load the data | ||
|
||
Load the | ||
[data from UCI repository](https://archive.ics.uci.edu/static/public/275/bike+sharing+dataset.zip) | ||
and save it locally. For demonstration purposes, we treat this data as the input | ||
data for a live model. To use with production models, you should make your | ||
prediction logs available. | ||
|
||
```cli | ||
$ wget https://archive.ics.uci.edu/static/public/275/bike+sharing+dataset.zip | ||
$ unzip bike+sharing+dataset.zip | ||
``` | ||
|
||
```python | ||
import pandas as pd | ||
|
||
df = pd.read_csv("raw_data/day.csv", header=0, sep=',', parse_dates=['dteday']) | ||
df.head() | ||
``` | ||
|
||
This is how it looks: ![Sample data](/img/evidently-sample-data-head.png) | ||
|
||
## Define column mapping | ||
|
||
You should specify the categorical and numerical features so that Evidently | ||
performs the correct statistical test for each of them. While Evidently can | ||
parse the data structure automatically, manually specifying the column type can | ||
minimize errors. | ||
|
||
```python | ||
from evidently.pipeline.column_mapping import ColumnMapping | ||
|
||
data_columns = ColumnMapping() | ||
data_columns.numerical_features = ['weathersit', 'temp', 'atemp', 'hum', 'windspeed'] | ||
data_columns.categorical_features = ['holiday', 'workingday'] | ||
``` | ||
|
||
## Define what to log | ||
|
||
Specify which metrics you want to calculate. In this case, you can generate the | ||
Data Drift report and log the drift score for each feature. | ||
|
||
```python | ||
from evidently.report import Report | ||
from evidently.metric_preset import DataDriftPreset | ||
|
||
|
||
def eval_drift(reference, production, column_mapping): | ||
data_drift_report = Report(metrics=[DataDriftPreset()]) | ||
data_drift_report.run( | ||
reference_data=reference, current_data=production, column_mapping=column_mapping | ||
) | ||
report = data_drift_report.as_dict() | ||
|
||
drifts = [] | ||
|
||
for feature in ( | ||
column_mapping.numerical_features + column_mapping.categorical_features | ||
): | ||
drifts.append( | ||
( | ||
feature, | ||
report["metrics"][1]["result"]["drift_by_columns"][feature][ | ||
"drift_score" | ||
], | ||
) | ||
) | ||
|
||
return drifts | ||
|
||
``` | ||
|
||
You can adapt what you want to calculate by selecting a different Preset or | ||
Metric from those available in Evidently. | ||
|
||
## Define the comparison windows | ||
|
||
Specify the period that is considered reference: Evidently will use it as the | ||
base for the comparison. Then, you should choose the periods to treat as | ||
experiments. This emulates the production model runs. | ||
|
||
```python | ||
#set reference dates | ||
reference_dates = ('2011-01-01 00:00:00','2011-01-28 23:00:00') | ||
|
||
#set experiment batches dates | ||
experiment_batches = [ | ||
('2011-01-01 00:00:00','2011-01-29 23:00:00'), | ||
('2011-01-29 00:00:00','2011-02-07 23:00:00'), | ||
('2011-02-07 00:00:00','2011-02-14 23:00:00'), | ||
('2011-02-15 00:00:00','2011-02-21 23:00:00'), | ||
] | ||
``` | ||
|
||
## Run and log experiments in DVC | ||
|
||
There are two ways to track the results of Evidently with DVCLive: | ||
|
||
1. you can save the results of each item in the batch in one single experiment | ||
(each experiment corresponds to a git commit), in separate steps (**our | ||
favorite option**) | ||
1. or you can save the result of each item in the batch as a separate experiment | ||
|
||
We will demonstrate both, and show you how to inspect the results regardless of | ||
your IDE. However, if you are using VSCode, we recommend using | ||
[our awesome VSCode extension](https://marketplace.visualstudio.com/items?itemName=Iterative.dvc) | ||
to inspect the results. | ||
|
||
### 1. One single experiment | ||
|
||
```python | ||
from dvclive import Live | ||
|
||
with Live() as live: | ||
for date in experiment_batches: | ||
live.log_param("begin", date[0]) | ||
live.log_param("end", date[1]) | ||
|
||
metrics = eval_drift( | ||
df.loc[df.dteday.between(reference_dates[0], reference_dates[1])], | ||
df.loc[df.dteday.between(date[0], date[1])], | ||
column_mapping=data_columns, | ||
) | ||
|
||
for feature in metrics: | ||
live.log_metric(feature[0], round(feature[1], 3)) | ||
|
||
live.next_step() | ||
``` | ||
|
||
You can then inspect the results using | ||
|
||
```cli | ||
$ dvc plots show | ||
``` | ||
|
||
and inspecting the resulting `dvc_plots/index.html`, which should look like | ||
this: ![DVC plot show](/img/evidently-dvc-plot-show.png) | ||
|
||
### 2. Multiple experiments | ||
|
||
```python | ||
from dvclive import Live | ||
|
||
for step, date in enumerate(experiment_batches): | ||
with Live() as live: | ||
live.log_param("begin", date[0]) | ||
live.log_param("end", date[1]) | ||
live.log_param("step", step) | ||
|
||
metrics = eval_drift( | ||
df.loc[df.dteday.between(reference_dates[0], reference_dates[1])], | ||
df.loc[df.dteday.between(date[0], date[1])], | ||
column_mapping=data_columns, | ||
) | ||
|
||
for feature in metrics: | ||
live.log_metric(feature[0], round(feature[1], 3)) | ||
|
||
``` | ||
|
||
You can the inspect the results using | ||
|
||
```cli | ||
$ dvc exp show | ||
``` | ||
|
||
``` | ||
ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ | ||
Experiment Created weathersit temp atemp hum windspeed holiday workingday step begin end | ||
ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ | ||
workspace - 0.231 0 0 0.062 0.012 0.275 0.593 3 2011-02-15 00:00:00 2011-02-21 23:00:00 | ||
master 10:02 AM - - - - - - - - - - | ||
βββ a96b45c [muggy-rand] 10:02 AM 0.231 0 0 0.062 0.012 0.275 0.593 3 2011-02-15 00:00:00 2011-02-21 23:00:00 | ||
βββ 78c6668 [pawky-arcs] 10:02 AM 0.155 0.399 0.537 0.684 0.611 0.588 0.699 2 2011-02-07 00:00:00 2011-02-14 23:00:00 | ||
βββ c1dd720 [joint-wont] 10:02 AM 0.779 0.098 0.107 0.03 0.171 0.545 0.653 1 2011-01-29 00:00:00 2011-02-07 23:00:00 | ||
βββ d0ddb8d [osmic-impi] 10:02 AM 0.985 1 1 1 1 0.98 0.851 0 2011-01-01 00:00:00 2011-01-29 23:00:00 | ||
ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ | ||
``` |
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.