pandasMFE 0.1.0__tar.gz

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
@@ -0,0 +1,21 @@
1
+ MIT License
2
+
3
+ Copyright (c) 2026 The pandasMFE authors
4
+
5
+ Permission is hereby granted, free of charge, to any person obtaining a copy
6
+ of this software and associated documentation files (the "Software"), to deal
7
+ in the Software without restriction, including without limitation the rights
8
+ to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
9
+ copies of the Software, and to permit persons to whom the Software is
10
+ furnished to do so, subject to the following conditions:
11
+
12
+ The above copyright notice and this permission notice shall be included in all
13
+ copies or substantial portions of the Software.
14
+
15
+ THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
16
+ IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
17
+ FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
18
+ AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
19
+ LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
20
+ OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
21
+ SOFTWARE.
@@ -0,0 +1,169 @@
1
+ Metadata-Version: 2.4
2
+ Name: pandasMFE
3
+ Version: 0.1.0
4
+ Summary: Risk and performance metrics (Sharpe, drawdown, VaR, CVaR) for pandas, with a .mfe accessor
5
+ Author: Gianmarco Scafarto
6
+ License-Expression: MIT
7
+ Keywords: pandas,finance,risk,var,cvar,sharpe,drawdown
8
+ Classifier: Programming Language :: Python :: 3
9
+ Classifier: Programming Language :: Python :: 3.11
10
+ Classifier: Programming Language :: Python :: 3.12
11
+ Classifier: Programming Language :: Python :: 3.13
12
+ Classifier: Operating System :: OS Independent
13
+ Classifier: Intended Audience :: Financial and Insurance Industry
14
+ Classifier: Topic :: Office/Business :: Financial :: Investment
15
+ Requires-Python: >=3.11
16
+ Description-Content-Type: text/markdown
17
+ License-File: LICENSE
18
+ Requires-Dist: pandas>=2.2
19
+ Requires-Dist: numpy>=1.26
20
+ Provides-Extra: excel
21
+ Requires-Dist: openpyxl>=3.1; extra == "excel"
22
+ Provides-Extra: data
23
+ Requires-Dist: yfinance; extra == "data"
24
+ Provides-Extra: dev
25
+ Requires-Dist: pytest>=7; extra == "dev"
26
+ Requires-Dist: build; extra == "dev"
27
+ Requires-Dist: twine; extra == "dev"
28
+ Requires-Dist: pyinstaller; extra == "dev"
29
+ Dynamic: license-file
30
+
31
+ # pandasMFE
32
+
33
+ Risk and performance metrics for **pandas**: annualized return (CAGR), volatility,
34
+ Sharpe ratio, maximum drawdown, historical and parametric VaR, CVaR (Expected Shortfall).
35
+
36
+ pandasMFE **extends pandas without modifying it**: after `import pandasmfe`, every
37
+ DataFrame and Series gains an `.mfe` namespace, using pandas' official
38
+ [accessor API](https://pandas.pydata.org/docs/development/extending.html).
39
+ Prices can be read from a CSV file or from a **SQLite database** with a SQL query.
40
+
41
+ > pandasMFE is an independent project and is not affiliated with the pandas project.
42
+
43
+ ## Installation
44
+
45
+ Requires Python 3.11+.
46
+
47
+ ```bash
48
+ pip install pandasMFE # library + `pandasmfe` command
49
+ pip install "pandasMFE[excel]" # + .xlsx export
50
+ pip install "pandasMFE[data]" # + yfinance, to download real prices
51
+ ```
52
+
53
+ ## Usage: the `.mfe` accessor
54
+
55
+ ```python
56
+ import pandas as pd
57
+ import pandasmfe # registers .mfe
58
+
59
+ prices = pd.read_csv("prices.csv", index_col=0, parse_dates=True)
60
+
61
+ # methods on PRICES
62
+ prices.mfe.risk_report(risk_free=0.02, alpha=0.95)
63
+ prices.mfe.max_drawdown()
64
+ prices.mfe.drawdown() # full drawdown series
65
+ returns = prices.mfe.returns()
66
+
67
+ # methods on RETURNS
68
+ returns.mfe.sharpe(risk_free=0.02)
69
+ returns.mfe.ann_volatility()
70
+ returns.mfe.var(alpha=0.99) # historical
71
+ returns.mfe.var(alpha=0.99, method="parametric") # Gaussian
72
+ returns.mfe.cvar(alpha=0.99)
73
+
74
+ # works on a single Series too, and chains
75
+ prices["STOCK_A"].mfe.returns().mfe.sharpe()
76
+ ```
77
+
78
+ ## Usage: plain functions
79
+
80
+ ```python
81
+ import pandasmfe as pm
82
+
83
+ pm.risk_report(prices, periods_per_year=252, risk_free=0.02, alpha=0.95)
84
+ pm.var_historical(pm.simple_returns(prices), alpha=0.99)
85
+ ```
86
+
87
+ ## Usage: real prices from a SQL database
88
+
89
+ Prices are stored in a SQLite database (one file, built into Python: no extra
90
+ dependency) and read back with a SQL query.
91
+
92
+ ```bash
93
+ pip install "pandasMFE[data]"
94
+ python examples/download_prices.py # SPY, AAPL, MSFT, TLT, GLD since 2015 -> examples/prices.db
95
+ ```
96
+
97
+ ```python
98
+ import pandasmfe as pm
99
+
100
+ prices = pm.read_prices_sql("examples/prices.db", tickers=["SPY", "TLT"], start="2020-01-01")
101
+ prices.mfe.risk_report(risk_free=0.02)
102
+
103
+ pm.list_tickers("examples/prices.db") # what is in the database
104
+ pm.write_prices_sql(my_prices, "my.db") # save any wide DataFrame of prices
105
+ ```
106
+
107
+ Table schema: `prices(date TEXT, ticker TEXT, close REAL)`, one row per date and ticker.
108
+ More queries in `examples/queries.sql`.
109
+
110
+ ## Usage: command line
111
+
112
+ ```bash
113
+ pandasmfe prices.csv
114
+ pandasmfe prices.csv --rf 0.02 --alpha 0.99 -o report.csv
115
+ pandasmfe excel_export.csv --eu -o report.xlsx # ';' separator, decimal comma, dd/mm/yyyy dates
116
+ pandasmfe examples/prices.db --tickers SPY TLT --start 2020-01-01
117
+ python -m pandasmfe prices.csv # equivalent
118
+ ```
119
+
120
+ `--tickers`, `--start` and `--end` work with both CSV files and databases.
121
+ Expected CSV format: first column = dates, then one price column per asset.
122
+
123
+ ```
124
+ Date,STOCK_A,STOCK_B
125
+ 2024-01-02,185.6,370.9
126
+ 2024-01-03,184.2,370.6
127
+ ```
128
+
129
+ `examples/sample_prices.csv` contains **synthetic** prices (geometric Brownian motion),
130
+ useful to try the package without downloading anything.
131
+
132
+ ## Standalone executable
133
+
134
+ The same command can be packaged with PyInstaller into a folder that runs on a
135
+ computer without Python:
136
+
137
+ ```bash
138
+ pip install pyinstaller
139
+ pyinstaller --onedir --name pandasmfe --clean --noconfirm standalone/pandasmfe_app.py
140
+ dist/pandasmfe/pandasmfe examples/sample_prices.csv
141
+ ```
142
+
143
+ PyInstaller builds for the operating system it runs on (a Windows `.exe` must be built on Windows).
144
+
145
+ ## Conventions
146
+
147
+ | Metric | Formula | Sign |
148
+ |---|---|---|
149
+ | `ann_return` | (∏(1+r))^(N/n) − 1 | ± |
150
+ | `ann_volatility` | σ · √N (ddof=1) | + |
151
+ | `sharpe` | mean(r − rf) / σ · √N | ± |
152
+ | `max_drawdown` | min(P / cummax(P) − 1) | ≤ 0 |
153
+ | `var_hist_95` | −quantile₅%(r) | positive loss |
154
+ | `var_param_95` | −(μ + z₅% σ) | positive loss |
155
+ | `cvar_hist_95` | −mean(r ∣ r ≤ quantile₅%) | positive loss |
156
+
157
+ N = periods per year (252 for daily data), n = number of observations.
158
+ Missing prices are never filled: a gap gives NaN returns, which are skipped.
159
+
160
+ ## Tests
161
+
162
+ ```bash
163
+ pip install -e ".[dev]"
164
+ pytest
165
+ ```
166
+
167
+ ## License
168
+
169
+ MIT
@@ -0,0 +1,139 @@
1
+ # pandasMFE
2
+
3
+ Risk and performance metrics for **pandas**: annualized return (CAGR), volatility,
4
+ Sharpe ratio, maximum drawdown, historical and parametric VaR, CVaR (Expected Shortfall).
5
+
6
+ pandasMFE **extends pandas without modifying it**: after `import pandasmfe`, every
7
+ DataFrame and Series gains an `.mfe` namespace, using pandas' official
8
+ [accessor API](https://pandas.pydata.org/docs/development/extending.html).
9
+ Prices can be read from a CSV file or from a **SQLite database** with a SQL query.
10
+
11
+ > pandasMFE is an independent project and is not affiliated with the pandas project.
12
+
13
+ ## Installation
14
+
15
+ Requires Python 3.11+.
16
+
17
+ ```bash
18
+ pip install pandasMFE # library + `pandasmfe` command
19
+ pip install "pandasMFE[excel]" # + .xlsx export
20
+ pip install "pandasMFE[data]" # + yfinance, to download real prices
21
+ ```
22
+
23
+ ## Usage: the `.mfe` accessor
24
+
25
+ ```python
26
+ import pandas as pd
27
+ import pandasmfe # registers .mfe
28
+
29
+ prices = pd.read_csv("prices.csv", index_col=0, parse_dates=True)
30
+
31
+ # methods on PRICES
32
+ prices.mfe.risk_report(risk_free=0.02, alpha=0.95)
33
+ prices.mfe.max_drawdown()
34
+ prices.mfe.drawdown() # full drawdown series
35
+ returns = prices.mfe.returns()
36
+
37
+ # methods on RETURNS
38
+ returns.mfe.sharpe(risk_free=0.02)
39
+ returns.mfe.ann_volatility()
40
+ returns.mfe.var(alpha=0.99) # historical
41
+ returns.mfe.var(alpha=0.99, method="parametric") # Gaussian
42
+ returns.mfe.cvar(alpha=0.99)
43
+
44
+ # works on a single Series too, and chains
45
+ prices["STOCK_A"].mfe.returns().mfe.sharpe()
46
+ ```
47
+
48
+ ## Usage: plain functions
49
+
50
+ ```python
51
+ import pandasmfe as pm
52
+
53
+ pm.risk_report(prices, periods_per_year=252, risk_free=0.02, alpha=0.95)
54
+ pm.var_historical(pm.simple_returns(prices), alpha=0.99)
55
+ ```
56
+
57
+ ## Usage: real prices from a SQL database
58
+
59
+ Prices are stored in a SQLite database (one file, built into Python: no extra
60
+ dependency) and read back with a SQL query.
61
+
62
+ ```bash
63
+ pip install "pandasMFE[data]"
64
+ python examples/download_prices.py # SPY, AAPL, MSFT, TLT, GLD since 2015 -> examples/prices.db
65
+ ```
66
+
67
+ ```python
68
+ import pandasmfe as pm
69
+
70
+ prices = pm.read_prices_sql("examples/prices.db", tickers=["SPY", "TLT"], start="2020-01-01")
71
+ prices.mfe.risk_report(risk_free=0.02)
72
+
73
+ pm.list_tickers("examples/prices.db") # what is in the database
74
+ pm.write_prices_sql(my_prices, "my.db") # save any wide DataFrame of prices
75
+ ```
76
+
77
+ Table schema: `prices(date TEXT, ticker TEXT, close REAL)`, one row per date and ticker.
78
+ More queries in `examples/queries.sql`.
79
+
80
+ ## Usage: command line
81
+
82
+ ```bash
83
+ pandasmfe prices.csv
84
+ pandasmfe prices.csv --rf 0.02 --alpha 0.99 -o report.csv
85
+ pandasmfe excel_export.csv --eu -o report.xlsx # ';' separator, decimal comma, dd/mm/yyyy dates
86
+ pandasmfe examples/prices.db --tickers SPY TLT --start 2020-01-01
87
+ python -m pandasmfe prices.csv # equivalent
88
+ ```
89
+
90
+ `--tickers`, `--start` and `--end` work with both CSV files and databases.
91
+ Expected CSV format: first column = dates, then one price column per asset.
92
+
93
+ ```
94
+ Date,STOCK_A,STOCK_B
95
+ 2024-01-02,185.6,370.9
96
+ 2024-01-03,184.2,370.6
97
+ ```
98
+
99
+ `examples/sample_prices.csv` contains **synthetic** prices (geometric Brownian motion),
100
+ useful to try the package without downloading anything.
101
+
102
+ ## Standalone executable
103
+
104
+ The same command can be packaged with PyInstaller into a folder that runs on a
105
+ computer without Python:
106
+
107
+ ```bash
108
+ pip install pyinstaller
109
+ pyinstaller --onedir --name pandasmfe --clean --noconfirm standalone/pandasmfe_app.py
110
+ dist/pandasmfe/pandasmfe examples/sample_prices.csv
111
+ ```
112
+
113
+ PyInstaller builds for the operating system it runs on (a Windows `.exe` must be built on Windows).
114
+
115
+ ## Conventions
116
+
117
+ | Metric | Formula | Sign |
118
+ |---|---|---|
119
+ | `ann_return` | (∏(1+r))^(N/n) − 1 | ± |
120
+ | `ann_volatility` | σ · √N (ddof=1) | + |
121
+ | `sharpe` | mean(r − rf) / σ · √N | ± |
122
+ | `max_drawdown` | min(P / cummax(P) − 1) | ≤ 0 |
123
+ | `var_hist_95` | −quantile₅%(r) | positive loss |
124
+ | `var_param_95` | −(μ + z₅% σ) | positive loss |
125
+ | `cvar_hist_95` | −mean(r ∣ r ≤ quantile₅%) | positive loss |
126
+
127
+ N = periods per year (252 for daily data), n = number of observations.
128
+ Missing prices are never filled: a gap gives NaN returns, which are skipped.
129
+
130
+ ## Tests
131
+
132
+ ```bash
133
+ pip install -e ".[dev]"
134
+ pytest
135
+ ```
136
+
137
+ ## License
138
+
139
+ MIT
@@ -0,0 +1,89 @@
1
+ # pyproject.toml — the package's configuration file.
2
+ # pip and `python -m build` read this file to know how to build the package,
3
+ # what it is called, which version it is and what it depends on.
4
+
5
+ # ---------------------------------------------------------------------------
6
+ # Which tool turns the source code into an installable package.
7
+ # ---------------------------------------------------------------------------
8
+ [build-system]
9
+ # setuptools does the building; version 77+ supports `license = "MIT"` (PEP 639).
10
+ requires = ["setuptools>=77"]
11
+ # The function inside setuptools that pip calls to build the package.
12
+ build-backend = "setuptools.build_meta"
13
+
14
+ # ---------------------------------------------------------------------------
15
+ # Package metadata: everything shown on PyPI and used by pip.
16
+ # ---------------------------------------------------------------------------
17
+ [project]
18
+ # Name on PyPI, used in `pip install pandasMFE` (must be unique on PyPI).
19
+ # The import name is different (`import pandasmfe`): it comes from the folder src/pandasmfe.
20
+ name = "pandasMFE"
21
+ # The version is not written here: setuptools reads it from src/pandasmfe/__init__.py
22
+ # (see [tool.setuptools.dynamic] below), so it lives in one single place.
23
+ dynamic = ["version"]
24
+ # One-line summary shown in PyPI search results.
25
+ description = "Risk and performance metrics (Sharpe, drawdown, VaR, CVaR) for pandas, with a .mfe accessor"
26
+ # This file becomes the project page on PyPI.
27
+ readme = "README.md"
28
+ # Older Pythons get a clear error from pip; 3.9 and 3.10 are end-of-life
29
+ # and pandas 3 requires Python 3.11+.
30
+ requires-python = ">=3.11"
31
+ # MIT licence: anyone may use, modify and share the code, keeping the copyright notice.
32
+ license = "MIT"
33
+ # The licence text is shipped inside the package.
34
+ license-files = ["LICENSE"]
35
+ # Authors shown on PyPI (add the other group members here if it is a group project).
36
+ authors = [{ name = "Gianmarco Scafarto" }]
37
+ # Free-text tags that help people find the package on PyPI.
38
+ keywords = ["pandas", "finance", "risk", "var", "cvar", "sharpe", "drawdown"]
39
+ # Standard labels from PyPI's official list (a misspelt classifier is rejected on upload).
40
+ classifiers = [
41
+ "Programming Language :: Python :: 3",
42
+ "Programming Language :: Python :: 3.11",
43
+ "Programming Language :: Python :: 3.12",
44
+ "Programming Language :: Python :: 3.13",
45
+ "Operating System :: OS Independent",
46
+
47
+ "Intended Audience :: Financial and Insurance Industry",
48
+ "Topic :: Office/Business :: Financial :: Investment",
49
+ ]
50
+ # Packages that pip installs automatically together with pandasMFE.
51
+ # Minimum versions (>=), never exact pins (==): a library must be installable
52
+ # next to other libraries. 2.2 and 1.26 are the oldest versions the tests were run on.
53
+ dependencies = [
54
+ "pandas>=2.2",
55
+ "numpy>=1.26",
56
+ ]
57
+
58
+ # Optional extras, installed only on request, e.g. pip install "pandasMFE[excel]".
59
+ [project.optional-dependencies]
60
+ # Needed only to save the report as an .xlsx file.
61
+ excel = ["openpyxl>=3.1"]
62
+ # Needed only by examples/download_prices.py to download real prices.
63
+ data = ["yfinance"]
64
+ # Tools for development: tests, building, uploading and the standalone executable.
65
+ dev = ["pytest>=7", "build", "twine", "pyinstaller"]
66
+
67
+ # Command-line programs created by pip on installation.
68
+ [project.scripts]
69
+ # Creates the `pandasmfe` command, which calls the function main() in pandasmfe/cli.py.
70
+ pandasmfe = "pandasmfe.cli:main"
71
+
72
+ # ---------------------------------------------------------------------------
73
+ # Settings for specific tools (each tool reads only its own section).
74
+ # ---------------------------------------------------------------------------
75
+ # Single source of the version: the variable __version__ in pandasmfe/__init__.py.
76
+ [tool.setuptools.dynamic]
77
+ version = { attr = "pandasmfe.__version__" }
78
+
79
+ # The package code lives in src/ ("src layout"): tests and examples stay out of the package.
80
+ [tool.setuptools.packages.find]
81
+ where = ["src"]
82
+
83
+ # Non-.py files to include in the package: the PEP 561 marker for type hints.
84
+ [tool.setuptools.package-data]
85
+ pandasmfe = ["py.typed"]
86
+
87
+ # pytest looks for tests in the tests/ folder.
88
+ [tool.pytest.ini_options]
89
+ testpaths = ["tests"]
@@ -0,0 +1,4 @@
1
+ [egg_info]
2
+ tag_build =
3
+ tag_date = 0
4
+
@@ -0,0 +1,169 @@
1
+ Metadata-Version: 2.4
2
+ Name: pandasMFE
3
+ Version: 0.1.0
4
+ Summary: Risk and performance metrics (Sharpe, drawdown, VaR, CVaR) for pandas, with a .mfe accessor
5
+ Author: Gianmarco Scafarto
6
+ License-Expression: MIT
7
+ Keywords: pandas,finance,risk,var,cvar,sharpe,drawdown
8
+ Classifier: Programming Language :: Python :: 3
9
+ Classifier: Programming Language :: Python :: 3.11
10
+ Classifier: Programming Language :: Python :: 3.12
11
+ Classifier: Programming Language :: Python :: 3.13
12
+ Classifier: Operating System :: OS Independent
13
+ Classifier: Intended Audience :: Financial and Insurance Industry
14
+ Classifier: Topic :: Office/Business :: Financial :: Investment
15
+ Requires-Python: >=3.11
16
+ Description-Content-Type: text/markdown
17
+ License-File: LICENSE
18
+ Requires-Dist: pandas>=2.2
19
+ Requires-Dist: numpy>=1.26
20
+ Provides-Extra: excel
21
+ Requires-Dist: openpyxl>=3.1; extra == "excel"
22
+ Provides-Extra: data
23
+ Requires-Dist: yfinance; extra == "data"
24
+ Provides-Extra: dev
25
+ Requires-Dist: pytest>=7; extra == "dev"
26
+ Requires-Dist: build; extra == "dev"
27
+ Requires-Dist: twine; extra == "dev"
28
+ Requires-Dist: pyinstaller; extra == "dev"
29
+ Dynamic: license-file
30
+
31
+ # pandasMFE
32
+
33
+ Risk and performance metrics for **pandas**: annualized return (CAGR), volatility,
34
+ Sharpe ratio, maximum drawdown, historical and parametric VaR, CVaR (Expected Shortfall).
35
+
36
+ pandasMFE **extends pandas without modifying it**: after `import pandasmfe`, every
37
+ DataFrame and Series gains an `.mfe` namespace, using pandas' official
38
+ [accessor API](https://pandas.pydata.org/docs/development/extending.html).
39
+ Prices can be read from a CSV file or from a **SQLite database** with a SQL query.
40
+
41
+ > pandasMFE is an independent project and is not affiliated with the pandas project.
42
+
43
+ ## Installation
44
+
45
+ Requires Python 3.11+.
46
+
47
+ ```bash
48
+ pip install pandasMFE # library + `pandasmfe` command
49
+ pip install "pandasMFE[excel]" # + .xlsx export
50
+ pip install "pandasMFE[data]" # + yfinance, to download real prices
51
+ ```
52
+
53
+ ## Usage: the `.mfe` accessor
54
+
55
+ ```python
56
+ import pandas as pd
57
+ import pandasmfe # registers .mfe
58
+
59
+ prices = pd.read_csv("prices.csv", index_col=0, parse_dates=True)
60
+
61
+ # methods on PRICES
62
+ prices.mfe.risk_report(risk_free=0.02, alpha=0.95)
63
+ prices.mfe.max_drawdown()
64
+ prices.mfe.drawdown() # full drawdown series
65
+ returns = prices.mfe.returns()
66
+
67
+ # methods on RETURNS
68
+ returns.mfe.sharpe(risk_free=0.02)
69
+ returns.mfe.ann_volatility()
70
+ returns.mfe.var(alpha=0.99) # historical
71
+ returns.mfe.var(alpha=0.99, method="parametric") # Gaussian
72
+ returns.mfe.cvar(alpha=0.99)
73
+
74
+ # works on a single Series too, and chains
75
+ prices["STOCK_A"].mfe.returns().mfe.sharpe()
76
+ ```
77
+
78
+ ## Usage: plain functions
79
+
80
+ ```python
81
+ import pandasmfe as pm
82
+
83
+ pm.risk_report(prices, periods_per_year=252, risk_free=0.02, alpha=0.95)
84
+ pm.var_historical(pm.simple_returns(prices), alpha=0.99)
85
+ ```
86
+
87
+ ## Usage: real prices from a SQL database
88
+
89
+ Prices are stored in a SQLite database (one file, built into Python: no extra
90
+ dependency) and read back with a SQL query.
91
+
92
+ ```bash
93
+ pip install "pandasMFE[data]"
94
+ python examples/download_prices.py # SPY, AAPL, MSFT, TLT, GLD since 2015 -> examples/prices.db
95
+ ```
96
+
97
+ ```python
98
+ import pandasmfe as pm
99
+
100
+ prices = pm.read_prices_sql("examples/prices.db", tickers=["SPY", "TLT"], start="2020-01-01")
101
+ prices.mfe.risk_report(risk_free=0.02)
102
+
103
+ pm.list_tickers("examples/prices.db") # what is in the database
104
+ pm.write_prices_sql(my_prices, "my.db") # save any wide DataFrame of prices
105
+ ```
106
+
107
+ Table schema: `prices(date TEXT, ticker TEXT, close REAL)`, one row per date and ticker.
108
+ More queries in `examples/queries.sql`.
109
+
110
+ ## Usage: command line
111
+
112
+ ```bash
113
+ pandasmfe prices.csv
114
+ pandasmfe prices.csv --rf 0.02 --alpha 0.99 -o report.csv
115
+ pandasmfe excel_export.csv --eu -o report.xlsx # ';' separator, decimal comma, dd/mm/yyyy dates
116
+ pandasmfe examples/prices.db --tickers SPY TLT --start 2020-01-01
117
+ python -m pandasmfe prices.csv # equivalent
118
+ ```
119
+
120
+ `--tickers`, `--start` and `--end` work with both CSV files and databases.
121
+ Expected CSV format: first column = dates, then one price column per asset.
122
+
123
+ ```
124
+ Date,STOCK_A,STOCK_B
125
+ 2024-01-02,185.6,370.9
126
+ 2024-01-03,184.2,370.6
127
+ ```
128
+
129
+ `examples/sample_prices.csv` contains **synthetic** prices (geometric Brownian motion),
130
+ useful to try the package without downloading anything.
131
+
132
+ ## Standalone executable
133
+
134
+ The same command can be packaged with PyInstaller into a folder that runs on a
135
+ computer without Python:
136
+
137
+ ```bash
138
+ pip install pyinstaller
139
+ pyinstaller --onedir --name pandasmfe --clean --noconfirm standalone/pandasmfe_app.py
140
+ dist/pandasmfe/pandasmfe examples/sample_prices.csv
141
+ ```
142
+
143
+ PyInstaller builds for the operating system it runs on (a Windows `.exe` must be built on Windows).
144
+
145
+ ## Conventions
146
+
147
+ | Metric | Formula | Sign |
148
+ |---|---|---|
149
+ | `ann_return` | (∏(1+r))^(N/n) − 1 | ± |
150
+ | `ann_volatility` | σ · √N (ddof=1) | + |
151
+ | `sharpe` | mean(r − rf) / σ · √N | ± |
152
+ | `max_drawdown` | min(P / cummax(P) − 1) | ≤ 0 |
153
+ | `var_hist_95` | −quantile₅%(r) | positive loss |
154
+ | `var_param_95` | −(μ + z₅% σ) | positive loss |
155
+ | `cvar_hist_95` | −mean(r ∣ r ≤ quantile₅%) | positive loss |
156
+
157
+ N = periods per year (252 for daily data), n = number of observations.
158
+ Missing prices are never filled: a gap gives NaN returns, which are skipped.
159
+
160
+ ## Tests
161
+
162
+ ```bash
163
+ pip install -e ".[dev]"
164
+ pytest
165
+ ```
166
+
167
+ ## License
168
+
169
+ MIT
@@ -0,0 +1,20 @@
1
+ LICENSE
2
+ README.md
3
+ pyproject.toml
4
+ src/pandasMFE.egg-info/PKG-INFO
5
+ src/pandasMFE.egg-info/SOURCES.txt
6
+ src/pandasMFE.egg-info/dependency_links.txt
7
+ src/pandasMFE.egg-info/entry_points.txt
8
+ src/pandasMFE.egg-info/requires.txt
9
+ src/pandasMFE.egg-info/top_level.txt
10
+ src/pandasmfe/__init__.py
11
+ src/pandasmfe/__main__.py
12
+ src/pandasmfe/accessor.py
13
+ src/pandasmfe/cli.py
14
+ src/pandasmfe/metrics.py
15
+ src/pandasmfe/py.typed
16
+ src/pandasmfe/sql.py
17
+ tests/test_accessor.py
18
+ tests/test_cli.py
19
+ tests/test_metrics.py
20
+ tests/test_sql.py
@@ -0,0 +1,2 @@
1
+ [console_scripts]
2
+ pandasmfe = pandasmfe.cli:main
@@ -0,0 +1,14 @@
1
+ pandas>=2.2
2
+ numpy>=1.26
3
+
4
+ [data]
5
+ yfinance
6
+
7
+ [dev]
8
+ pytest>=7
9
+ build
10
+ twine
11
+ pyinstaller
12
+
13
+ [excel]
14
+ openpyxl>=3.1
@@ -0,0 +1 @@
1
+ pandasmfe
@@ -0,0 +1,55 @@
1
+ """pandasMFE — risk and performance metrics for pandas, with a ``.mfe`` accessor.
2
+
3
+ Two ways to use it::
4
+
5
+ import pandasmfe as pm
6
+ pm.risk_report(prices) # plain functions
7
+
8
+ prices.mfe.risk_report() # accessor, available after `import pandasmfe`
9
+
10
+ Prices can come from a CSV or from a SQLite database (``pm.read_prices_sql``).
11
+ """
12
+
13
+ # Importing accessor.py runs its decorators, which register `.mfe` on every
14
+ # DataFrame and Series. The import is needed for this side effect only, so the
15
+ # name is never used ("noqa: F401" tells linters the unused import is intentional).
16
+ from pandasmfe import accessor as _accessor # noqa: F401
17
+
18
+ # Re-export the public functions, so users write pm.risk_report(...)
19
+ # instead of pm.metrics.risk_report(...). Same objects, not copies.
20
+ from pandasmfe.metrics import (
21
+ annualized_return,
22
+ annualized_volatility,
23
+ cvar_historical,
24
+ drawdown_series,
25
+ max_drawdown,
26
+ risk_report,
27
+ sharpe_ratio,
28
+ simple_returns,
29
+ var_historical,
30
+ var_parametric,
31
+ )
32
+ # Re-export the database functions in the same way.
33
+ from pandasmfe.sql import list_tickers, read_prices_sql, write_prices_sql
34
+
35
+ # The package version, written in this single place; pyproject.toml reads it from here.
36
+ __version__ = "0.1.0"
37
+
38
+ # The official list of public names: what `from pandasmfe import *` imports,
39
+ # and what users can rely on. Internal modules and helpers are left out.
40
+ __all__ = [
41
+ "simple_returns",
42
+ "annualized_return",
43
+ "annualized_volatility",
44
+ "sharpe_ratio",
45
+ "drawdown_series",
46
+ "max_drawdown",
47
+ "var_historical",
48
+ "var_parametric",
49
+ "cvar_historical",
50
+ "risk_report",
51
+ "read_prices_sql",
52
+ "write_prices_sql",
53
+ "list_tickers",
54
+ "__version__",
55
+ ]