rosie-df 0.1.0__tar.gz

rosie_df-0.1.0/PKG-INFO

@@ -0,0 +1,11 @@
+Metadata-Version: 2.4
+Name: rosie-df
+Version: 0.1.0
+Summary: High-performance DataFrame Rosetta Stone powered by Polars
+Author-email: anandlenin <anandijjina719@gmail.com>
+Requires-Python: >=3.9
+Description-Content-Type: text/markdown
+Requires-Dist: pyyaml>=6.0
+Requires-Dist: polars>=1.0.0
+
+Cheatsheet tool mapping syntax across Pandas, Polars, and PySpark.

rosie_df-0.1.0/README.md

@@ -0,0 +1 @@
+Cheatsheet tool mapping syntax across Pandas, Polars, and PySpark.

rosie_df-0.1.0/pyproject.toml

@@ -0,0 +1,23 @@
+[build-system]
+requires = ["setuptools>=61.0"]
+build-backend = "setuptools.build_meta"
+
+[project]
+name = "rosie-df"
+version = "0.1.0"
+description = "High-performance DataFrame Rosetta Stone powered by Polars"
+readme = "README.md"
+requires-python = ">=3.9"
+authors = [
+    { name = "anandlenin", email = "anandijjina719@gmail.com" }
+]
+dependencies = [
+    "pyyaml>=6.0",
+    "polars>=1.0.0"
+]
+
+[project.scripts]
+rosie = "rosie_df.cli:main"
+
+[tool.setuptools.package-data]
+rosie_df = ["data/*.yaml"]

rosie_df-0.1.0/rosie_df/__init__.py

@@ -0,0 +1,3 @@
+from rosie_df.core import helper, show, SyntaxLookup
+
+__all__ = ["helper", "show", "SyntaxLookup"]

rosie_df-0.1.0/rosie_df/cli.py

@@ -0,0 +1,39 @@
+import argparse
+import polars as pl
+from rosie_df.core import helper
+
+def main():
+    parser = argparse.ArgumentParser(
+        prog="rosie",
+        description="Rosie-DF: Fast DataFrame Cheatsheet for Pandas, Polars & PySpark"
+    )
+    parser.add_argument("engines", nargs="*", default=[], help="Engines: pandas, polars, pyspark")
+    parser.add_argument("--category", "-c", type=str, default=None, help="Filter category: inspect, clean, transform, eda, export")
+    parser.add_argument("--search", "-s", type=str, default=None, help="Search query")
+    parser.add_argument("--ops", "-o", nargs="+", default=None, help="Filter specific operations")
+    parser.add_argument("--gotchas", "-g", action="store_true", help="Show performance gotchas")
+    parser.add_argument("--transpose", "-t", action="store_true", help="Transpose rows and columns")
+    parser.add_argument("--categories", action="store_true", help="List available categories")
+
+    args = parser.parse_args()
+
+    if args.categories:
+        print("\nAvailable Categories:", ", ".join(helper.categories()), "\n")
+        return
+
+    if args.search:
+        result = helper.search(args.search, *args.engines, gotchas=args.gotchas)
+    else:
+        result = helper.show(
+            *args.engines,
+            category=args.category,
+            ops=args.ops,
+            gotchas=args.gotchas,
+            transpose=True if args.transpose else None
+        )
+
+    with pl.Config(tbl_rows=-1, tbl_cols=-1, fmt_str_lengths=120, tbl_width_chars=180):
+        print("\n", result.df, "\n")
+
+if __name__ == "__main__":
+    main()

rosie_df-0.1.0/rosie_df/core.py

@@ -0,0 +1,212 @@
+import html
+from pathlib import Path
+from typing import List, Optional
+import yaml
+import polars as pl
+
+YAML_FILE = Path(__file__).parent / "data" / "cheatsheet.yaml"
+VALID_ENGINES = ["pandas", "polars", "pyspark"]
+
+
+class RosieResult:
+    def __init__(self, df: pl.DataFrame, raw_data: list, engines: list, gotchas: bool, transpose: bool):
+        self.df = df
+        self.raw_data = raw_data
+        self.engines = engines
+        self.gotchas = gotchas
+        self.transpose = transpose
+
+    def _repr_html_(self) -> str:
+        css = """
+        <style>
+            .rosie-table {
+                width: 100%;
+                border-collapse: collapse;
+                font-family: -apple-system, BlinkMacSystemFont, "Segoe UI", Roboto, Helvetica, Arial, sans-serif;
+                margin: 12px 0;
+                font-size: 13px;
+                line-height: 1.4;
+            }
+            .rosie-table th {
+                background: rgba(125, 125, 125, 0.15);
+                color: var(--jp-ui-font-color1, #e0e0e0);
+                font-weight: 600;
+                padding: 10px 14px;
+                border: 1px solid rgba(125, 125, 125, 0.25);
+                text-align: left;
+            }
+            .rosie-table td {
+                padding: 10px 14px;
+                border: 1px solid rgba(125, 125, 125, 0.25);
+                vertical-align: top;
+            }
+            .rosie-badge {
+                display: inline-block;
+                padding: 3px 8px;
+                border-radius: 4px;
+                font-weight: bold;
+                font-size: 11px;
+                text-transform: uppercase;
+            }
+            .badge-pandas { background: #306998; color: #ffffff; }
+            .badge-polars { background: #00876c; color: #ffffff; }
+            .badge-pyspark { background: #e25a1c; color: #ffffff; }
+            .rosie-code {
+                font-family: "JetBrains Mono", Consolas, Menlo, monospace;
+                font-size: 12px;
+                background: rgba(125, 125, 125, 0.1);
+                border: 1px solid rgba(125, 125, 125, 0.2);
+                border-radius: 4px;
+                padding: 8px 10px;
+                margin: 0;
+                white-space: pre-wrap;
+                word-break: break-word;
+                color: var(--jp-ui-font-color1, #ffffff);
+            }
+            .rosie-gotcha {
+                margin-top: 6px;
+                padding: 6px 8px;
+                background: rgba(245, 158, 11, 0.12);
+                border-left: 3px solid #f59e0b;
+                border-radius: 2px;
+                font-size: 11px;
+                color: var(--jp-ui-font-color2, #f3f4f6);
+            }
+        </style>
+        """
+
+        html_out = [css, '<table class="rosie-table">']
+
+        if not self.transpose:
+            html_out.append("<thead><tr><th>Library</th>")
+            for item in self.raw_data:
+                op_name = item.get("name", item.get("id"))
+                html_out.append(f"<th>{html.escape(op_name)}</th>")
+            html_out.append("</tr></thead><tbody>")
+
+            for eng in self.engines:
+                html_out.append("<tr>")
+                html_out.append(f'<td><span class="rosie-badge badge-{eng}">{eng}</span></td>')
+                for item in self.raw_data:
+                    syntax = item.get(eng, "").strip()
+                    code_html = f'<pre class="rosie-code">{html.escape(syntax)}</pre>'
+                    gotcha_html = ""
+                    if self.gotchas:
+                        note = item.get("gotchas", {}).get(eng, "")
+                        if note:
+                            gotcha_html = f'<div class="rosie-gotcha">💡 <b>Gotcha:</b> {html.escape(note)}</div>'
+                    html_out.append(f"<td>{code_html}{gotcha_html}</td>")
+                html_out.append("</tr>")
+        else:
+            html_out.append("<thead><tr><th>Operation</th>")
+            for eng in self.engines:
+                html_out.append(f'<th><span class="rosie-badge badge-{eng}">{eng}</span></th>')
+            html_out.append("</tr></thead><tbody>")
+
+            for item in self.raw_data:
+                op_name = item.get("name", item.get("id"))
+                html_out.append(f"<tr><td><b>{html.escape(op_name)}</b></td>")
+                for eng in self.engines:
+                    syntax = item.get(eng, "").strip()
+                    code_html = f'<pre class="rosie-code">{html.escape(syntax)}</pre>'
+                    gotcha_html = ""
+                    if self.gotchas:
+                        note = item.get("gotchas", {}).get(eng, "")
+                        if note:
+                            gotcha_html = f'<div class="rosie-gotcha">💡 <b>Gotcha:</b> {html.escape(note)}</div>'
+                    html_out.append(f"<td>{code_html}{gotcha_html}</td>")
+                html_out.append("</tr>")
+
+        html_out.append("</tbody></table>")
+        return "".join(html_out)
+
+    def to_polars(self) -> pl.DataFrame:
+        return self.df
+
+    def __repr__(self) -> str:
+        with pl.Config(tbl_rows=-1, tbl_cols=-1, fmt_str_lengths=120, tbl_width_chars=180):
+            return str(self.df)
+
+
+class SyntaxLookup:
+    def __init__(self, data_path: Path = YAML_FILE):
+        self.data_path = data_path
+        self._raw_data = self._load_data()
+
+    def _load_data(self) -> list:
+        if not self.data_path.exists():
+            raise FileNotFoundError(f"Cheatsheet data not found at: {self.data_path}")
+        with open(self.data_path, "r", encoding="utf-8") as f:
+            return yaml.safe_load(f).get("operations", [])
+
+    def categories(self) -> List[str]:
+        return sorted(list(set(item.get("category", "general") for item in self._raw_data)))
+
+    def show(
+        self,
+        *engines: str,
+        category: Optional[str] = None,
+        ops: Optional[List[str]] = None,
+        gotchas: bool = False,
+        transpose: Optional[bool] = None
+    ) -> RosieResult:
+        """
+        Compare DataFrame syntax across engines.
+
+        Categories:
+            - 'inspect'   : read_data, view_rows, schema_info, count_nulls, drop_duplicates, describe_stats
+            - 'clean'     : filter_rows, handle_nulls, cast_types, rename_columns, string_cleaning
+            - 'transform' : add_column, groupby_agg, join_tables, reshape, window_functions
+            - 'eda'       : univariate_counts, quantiles, skew_kurt, bivariate_corr, crosstab, covariance
+            - 'export'    : save_parquet, save_csv
+        """
+        selected_engines = [e.lower() for e in engines] if engines else VALID_ENGINES
+        for eng in selected_engines:
+            if eng not in VALID_ENGINES:
+                raise ValueError(f"Unknown engine '{eng}'. Choose from: {VALID_ENGINES}")
+
+        filtered_items = []
+        for item in self._raw_data:
+            if category and item.get("category", "").lower() != category.lower():
+                continue
+            op_id = item.get("id", "")
+            op_name = item.get("name", op_id)
+            if ops and not any(t.lower() in (op_id.lower(), op_name.lower()) for t in ops):
+                continue
+            filtered_items.append(item)
+
+        # Auto-transpose if table is wide (> 3 columns) unless explicitly overridden
+        should_transpose = transpose if transpose is not None else (len(filtered_items) > 3)
+
+        if not should_transpose:
+            data_dict = {"Library": selected_engines}
+            for item in filtered_items:
+                name = item.get("name", item.get("id"))
+                data_dict[name] = [item.get(eng, "").strip() for eng in selected_engines]
+            df = pl.DataFrame(data_dict)
+        else:
+            data_dict = {"Operation": [item.get("name", item.get("id")) for item in filtered_items]}
+            for eng in selected_engines:
+                data_dict[eng] = [item.get(eng, "").strip() for item in filtered_items]
+            df = pl.DataFrame(data_dict)
+
+        return RosieResult(df, filtered_items, selected_engines, gotchas, should_transpose)
+
+    def search(self, query: str, *engines: str, gotchas: bool = False) -> RosieResult:
+        """Search query matching across name, id, category, and gotchas."""
+        q = query.lower()
+        matched_ops = []
+        for item in self._raw_data:
+            fields = [
+                item.get("id", ""),
+                item.get("name", ""),
+                item.get("category", ""),
+                str(item.get("gotchas", ""))
+            ]
+            if any(q in f.lower() for f in fields):
+                matched_ops.append(item.get("id"))
+        return self.show(*engines, ops=matched_ops, gotchas=gotchas)
+
+
+helper = SyntaxLookup()
+show = helper.show

rosie_df-0.1.0/rosie_df/data/cheatsheet.yaml

@@ -0,0 +1,279 @@
+operations:
+  # ==========================================
+  # 1. INSPECT (category: "inspect")
+  # ==========================================
+  - id: read_data
+    name: "Read Data (CSV)"
+    category: "inspect"
+    pandas: "pd.read_csv('file.csv')"
+    polars: "pl.read_csv('file.csv')\n# or pl.scan_csv('file.csv') for lazy"
+    pyspark: "spark.read.csv('file.csv', header=True, inferSchema=True)"
+    gotchas:
+      pandas: "Eager only; uses 5-10x memory of raw file size."
+      polars: "Use scan_csv() to optimize execution plans before loading."
+      pyspark: "inferSchema=True triggers two full reads of the dataset."
+
+  - id: view_rows
+    name: "Preview Rows (head)"
+    category: "inspect"
+    pandas: "df.head(5)"
+    polars: "df.head(5)"
+    pyspark: "df.show(5)"
+    gotchas:
+      pandas: "Returns a new DataFrame slice."
+      polars: "Runs without copying memory."
+      pyspark: "df.head(5) returns Row objects; use df.show(5) for tabular view."
+
+  - id: schema_info
+    name: "Schema & Types (info)"
+    category: "inspect"
+    pandas: "df.info()"
+    polars: "df.schema\n# or df.glimpse()"
+    pyspark: "df.printSchema()"
+    gotchas:
+      pandas: "df.info() prints directly to stdout; does not return an object."
+      polars: "No df.info(). Use df.glimpse() for compact single-screen view."
+      pyspark: "df.printSchema() prints tree structure to console."
+
+  - id: count_nulls
+    name: "Count Nulls"
+    category: "inspect"
+    pandas: "df.isnull().sum()"
+    polars: "df.null_count()"
+    pyspark: "df.select([F.count(F.when(F.col(c).isNull(), c)).alias(c) for c in df.columns])"
+    gotchas:
+      pandas: "Calculated column by column eagerly."
+      polars: "null_count() returns a 1-row DataFrame multithreaded."
+      pyspark: "Spark has no built-in null_count(); requires list comprehension."
+
+  - id: drop_duplicates
+    name: "Deduplication"
+    category: "inspect"
+    pandas: "df.drop_duplicates(subset=['id'])"
+    polars: "df.unique(subset=['id'])"
+    pyspark: "df.dropDuplicates(['id'])"
+    gotchas:
+      pandas: "Defaults to keep='first'."
+      polars: "Method is unique(), not drop_duplicates()."
+      pyspark: "Triggers global shuffle across nodes."
+
+  - id: describe_stats
+    name: "Summary Statistics"
+    category: "inspect"
+    pandas: "df.describe()"
+    polars: "df.describe()"
+    pyspark: "df.describe().show()"
+    gotchas:
+      pandas: "Includes percentiles 25, 50, 75 by default."
+      polars: "Returns a DataFrame with min, max, mean, std, null_count."
+      pyspark: "Computes count, mean, stddev, min, max eagerly."
+
+  # ==========================================
+  # 2. CLEAN (category: "clean")
+  # ==========================================
+  - id: filter_rows
+    name: "Filter Rows"
+    category: "clean"
+    pandas: "df[(df['age'] > 25) & (df['status'] == 'active')]"
+    polars: "df.filter((pl.col('age') > 25) & (pl.col('status') == 'active'))"
+    pyspark: "df.filter((F.col('age') > 25) & (F.col('status') == 'active'))"
+    gotchas:
+      pandas: "Missing parentheses around conditions causes bitwise precedence errors."
+      polars: "Never use Python and/or; use & and |."
+      pyspark: "SQL string syntax also allowed: df.filter('age > 25 AND status = \"active\"')."
+
+  - id: handle_nulls
+    name: "Drop or Fill Nulls"
+    category: "clean"
+    pandas: "df.dropna(subset=['col'])\ndf['col'].fillna(0)"
+    polars: "df.drop_nulls('col')\ndf.with_columns(pl.col('col').fill_null(0))"
+    pyspark: "df.dropna(subset=['col'])\ndf.fillna({'col': 0})"
+    gotchas:
+      pandas: "fillna in place modifies state; returns None if inplace=True."
+      polars: "fill_null() also supports strategies: 'forward', 'backward', 'mean'."
+      pyspark: "df.fillna() converts numeric targets automatically."
+
+  - id: cast_types
+    name: "Cast Data Types"
+    category: "clean"
+    pandas: "df['age'] = df['age'].astype('int64')"
+    polars: "df = df.with_columns(pl.col('age').cast(pl.Int64))"
+    pyspark: "df = df.withColumn('age', F.col('age').cast('long'))"
+    gotchas:
+      pandas: "Astype fails completely if column contains NaN for integer casts."
+      polars: "Use strict=False in cast() to turn invalid parses into nulls without error."
+      pyspark: "Invalid casts silently return null in Spark."
+
+  - id: rename_columns
+    name: "Rename Columns"
+    category: "clean"
+    pandas: "df.rename(columns={'old_name': 'new_name'})"
+    polars: "df.rename({'old_name': 'new_name'})"
+    pyspark: "df.withColumnRenamed('old_name', 'new_name')"
+    gotchas:
+      pandas: "Requires columns={} argument explicitly."
+      polars: "Renaming is zero-copy metadata update."
+      pyspark: "Chain withColumnRenamed multiple times or use select alias."
+
+  - id: string_cleaning
+    name: "String Cleaning (Trim / Lower)"
+    category: "clean"
+    pandas: "df['name'].str.strip().str.lower()"
+    polars: "df.with_columns(pl.col('name').str.strip_chars().str.to_lowercase())"
+    pyspark: "df.withColumn('name', F.lower(F.trim(F.col('name'))))"
+    gotchas:
+      pandas: "Accessor is .str; can be slow on large string series."
+      polars: "strip_chars() removes whitespace; strip_prefix() for prefixes."
+      pyspark: "Requires functions imported from pyspark.sql.functions as F."
+
+  # ==========================================
+  # 3. TRANSFORM (category: "transform")
+  # ==========================================
+  - id: add_column
+    name: "Add / Mutate Column"
+    category: "transform"
+    pandas: "df['bonus'] = df['salary'] * 0.1"
+    polars: "df.with_columns((pl.col('salary') * 0.1).alias('bonus'))"
+    pyspark: "df.withColumn('bonus', F.col('salary') * 0.1)"
+    gotchas:
+      pandas: "Can throw SettingWithCopyWarning on dataframe slices."
+      polars: "Place multiple expressions inside one with_columns() call for concurrency."
+      pyspark: "Chaining many withColumn calls degrades Catalyst optimizer plan."
+
+  - id: groupby_agg
+    name: "GroupBy & Aggregate"
+    category: "transform"
+    pandas: "df.groupby('dept').agg(avg_sal=('salary', 'mean')).reset_index()"
+    polars: "df.group_by('dept').agg(pl.col('salary').mean().alias('avg_sal'))"
+    pyspark: "df.groupBy('dept').agg(F.mean('salary').alias('avg_sal'))"
+    gotchas:
+      pandas: "Pandas creates index on group keys; must call reset_index()."
+      polars: "Ordering not preserved unless explicitly called with sort()."
+      pyspark: "Triggers wide network shuffle."
+
+  - id: join_tables
+    name: "Join Tables"
+    category: "transform"
+    pandas: "pd.merge(df1, df2, on='id', how='left')"
+    polars: "df1.join(df2, on='id', how='left')"
+    pyspark: "df1.join(df2, on='id', how='left')"
+    gotchas:
+      pandas: "Different keys use left_on='a', right_on='b'."
+      polars: "Different keys use left_on='a', right_on='b'."
+      pyspark: "Duplicate join columns remain in schema unless dropped or joined with array."
+
+  - id: reshape
+    name: "Reshape (Pivot / Melt)"
+    category: "transform"
+    pandas: "# Pivot wide:\ndf.pivot(index='id', columns='var', values='val')\n# Melt long:\ndf.melt(id_vars=['id'], value_vars=['a', 'b'])"
+    polars: "# Pivot wide:\ndf.pivot(on='var', index='id', values='val')\n# Melt long:\ndf.unpivot(index=['id'], on=['a', 'b'])"
+    pyspark: "# Pivot wide:\ndf.groupBy('id').pivot('var').sum('val')\n# Melt long:\ndf.selectExpr('id', 'stack(2, \"a\", a, \"b\", b) as (var, val)')"
+    gotchas:
+      pandas: "Pandas uses melt(); Polars 1.0+ renamed melt() to unpivot()."
+      polars: "Pivot requires eager DataFrame (not LazyFrame)."
+      pyspark: "Unpivot in Spark uses stack() expression via selectExpr."
+
+  - id: window_functions
+    name: "Window Functions (Rank)"
+    category: "transform"
+    pandas: "df['rank'] = df.groupby('dept')['salary'].rank(ascending=False)"
+    polars: "df.with_columns(pl.col('salary').rank(descending=True).over('dept').alias('rank'))"
+    pyspark: "w = Window.partitionBy('dept').orderBy(F.col('salary').desc())\ndf.withColumn('rank', F.rank().over(w))"
+    gotchas:
+      pandas: "Requires groupby followed by transform or rank."
+      polars: ".over() replaces complex joins without memory penalty."
+      pyspark: "Omitting partitionBy() loads all data into single node causing OOM."
+
+  # ==========================================
+  # 4. EDA (category: "eda")
+  # ==========================================
+  - id: univariate_counts
+    name: "Value Counts (Univariate)"
+    category: "eda"
+    pandas: "df['category'].value_counts(normalize=True)"
+    polars: "df['category'].value_counts(normalize=True)"
+    pyspark: "df.groupBy('category').count().withColumn('pct', F.col('count') / df.count())"
+    gotchas:
+      pandas: "Direct Series method."
+      polars: "Returns a DataFrame with columns [category, count]."
+      pyspark: "No native normalize param; calculate count / total."
+
+  - id: univariate_quantiles
+    name: "Quantiles / Percentiles (Univariate)"
+    category: "eda"
+    pandas: "df['age'].quantile([0.25, 0.5, 0.75])"
+    polars: "df.select([pl.col('age').quantile(0.5)])"
+    pyspark: "df.approxQuantile('age', [0.25, 0.5, 0.75], 0.01)"
+    gotchas:
+      pandas: "Calculates exact quantiles in-memory."
+      polars: "Supports exact quantile on Column/Series."
+      pyspark: "Uses Greenwald-Khanna approx algorithm; 0.01 is relative error."
+
+  - id: univariate_skew_kurt
+    name: "Skewness & Kurtosis (Univariate)"
+    category: "eda"
+    pandas: "skew = df['val'].skew()\nkurt = df['val'].kurt()"
+    polars: "df.select([pl.col('val').skew().alias('skew'), pl.col('val').kurtosis().alias('kurt')])"
+    pyspark: "df.select(F.skewness('val').alias('skew'), F.kurtosis('val').alias('kurt'))"
+    gotchas:
+      pandas: "Pandas calculates unbiased Fisher-Pearson kurtosis."
+      polars: "Zero-copy expression execution inside select."
+      pyspark: "Distributed computation via built-in functions."
+
+  - id: bivariate_corr
+    name: "Correlation Matrix (Bivariate)"
+    category: "eda"
+    pandas: "df[['col_a', 'col_b']].corr(method='pearson')"
+    polars: "df.select(pl.corr('col_a', 'col_b', method='pearson'))"
+    pyspark: "df.stat.corr('col_a', 'col_b', method='pearson')"
+    gotchas:
+      pandas: "corr() on full dataframe creates N x N matrix."
+      polars: "pl.corr() is an expression calculated between two columns."
+      pyspark: "stat.corr() works pairwise between two columns only."
+
+  - id: bivariate_crosstab
+    name: "Cross-Tabulation (Bivariate)"
+    category: "eda"
+    pandas: "pd.crosstab(df['cat_a'], df['cat_b'])"
+    polars: "df.pivot(on='cat_b', index='cat_a', values='cat_a', aggregate_function='len')"
+    pyspark: "df.stat.crosstab('cat_a', 'cat_b')"
+    gotchas:
+      pandas: "pd.crosstab is standalone function taking Series."
+      polars: "Polars builds cross-tab using pivot with aggregate_function='len'."
+      pyspark: "Native stat.crosstab returns a DataFrame."
+
+  - id: multivariate_cov
+    name: "Covariance (Multivariate)"
+    category: "eda"
+    pandas: "df[['a', 'b', 'c']].cov()"
+    polars: "df.select(pl.cov('a', 'b'))"
+    pyspark: "df.stat.cov('a', 'b')"
+    gotchas:
+      pandas: "Calculates full N x N covariance matrix."
+      polars: "Expression-based; compute pairs via pl.cov(col1, col2)."
+      pyspark: "stat.cov(col1, col2) calculates scalar covariance."
+
+  # ==========================================
+  # 5. EXPORT (category: "export")
+  # ==========================================
+  - id: save_parquet
+    name: "Export Parquet"
+    category: "export"
+    pandas: "df.to_parquet('output.parquet', compression='snappy')"
+    polars: "df.write_parquet('output.parquet', compression='snappy')"
+    pyspark: "df.write.mode('overwrite').parquet('output.parquet')"
+    gotchas:
+      pandas: "Requires pyarrow or fastparquet installed."
+      polars: "Native Rust multi-threaded writer with zstd/snappy."
+      pyspark: "Writes directory of partition files, not single file."
+
+  - id: save_csv
+    name: "Export CSV"
+    category: "export"
+    pandas: "df.to_csv('output.csv', index=False)"
+    polars: "df.write_csv('output.csv')"
+    pyspark: "df.write.mode('overwrite').csv('output.csv', header=True)"
+    gotchas:
+      pandas: "Always set index=False or it prepends extra unnamed column."
+      polars: "Never outputs row indexes."
+      pyspark: "Outputs part-* files in folder unless coalesced."

rosie_df-0.1.0/rosie_df.egg-info/PKG-INFO

@@ -0,0 +1,11 @@
+Metadata-Version: 2.4
+Name: rosie-df
+Version: 0.1.0
+Summary: High-performance DataFrame Rosetta Stone powered by Polars
+Author-email: anandlenin <anandijjina719@gmail.com>
+Requires-Python: >=3.9
+Description-Content-Type: text/markdown
+Requires-Dist: pyyaml>=6.0
+Requires-Dist: polars>=1.0.0
+
+Cheatsheet tool mapping syntax across Pandas, Polars, and PySpark.

rosie_df-0.1.0/rosie_df.egg-info/SOURCES.txt

@@ -0,0 +1,12 @@
+README.md
+pyproject.toml
+rosie_df/__init__.py
+rosie_df/cli.py
+rosie_df/core.py
+rosie_df.egg-info/PKG-INFO
+rosie_df.egg-info/SOURCES.txt
+rosie_df.egg-info/dependency_links.txt
+rosie_df.egg-info/entry_points.txt
+rosie_df.egg-info/requires.txt
+rosie_df.egg-info/top_level.txt
+rosie_df/data/cheatsheet.yaml

rosie_df-0.1.0/rosie_df.egg-info/dependency_links.txt

@@ -0,0 +1 @@
+

rosie_df-0.1.0/rosie_df.egg-info/entry_points.txt

@@ -0,0 +1,2 @@
+[console_scripts]
+rosie = rosie_df.cli:main

rosie_df-0.1.0/rosie_df.egg-info/requires.txt

@@ -0,0 +1,2 @@
+pyyaml>=6.0
+polars>=1.0.0

rosie_df-0.1.0/rosie_df.egg-info/top_level.txt

@@ -0,0 +1 @@
+rosie_df

rosie_df-0.1.0/setup.cfg

@@ -0,0 +1,4 @@
+[egg_info]
+tag_build = 
+tag_date = 0
+