20. Glossary
Ha glossary — Data Science terms with English meaning aani Marathi arth. Rojche shabd (filter, plot, notebook) English madhech vapra. Visarla tar ithe bagha.
| Term | Meaning | मराठी अर्थ |
|---|---|---|
| Aggregation | Summarising many values into one (sum, mean, count). | एकत्रीकरण |
| Axes (matplotlib) | The plotting area object that holds titles, labels and artists. | — (English madhech vapra) |
| Boolean mask | True/False Series used to filter rows. | खरे/खोटे यानुसार ओळी निवडणे |
| Box plot | Chart showing median, quartiles and outliers. | — (English madhech vapra) |
| CSV | Comma-separated values text file for tables. | — (English madhech vapra) |
| DataFrame | pandas table with labelled columns and an index. | — (English madhech vapra) |
| dtype | Data type of a column or array (int64, float64, object…). | डेटा प्रकार |
| EDA | Exploratory Data Analysis — structured look at a dataset before reporting or modelling. | शोधक डेटा विश्लेषण |
| Figure | The whole matplotlib window/canvas that can hold one or more axes. | — (English madhech vapra) |
| Grain | What one row represents (order line, order, day×city…). | तपशीलाची पातळी |
| GroupBy | Split–apply–combine grouping in pandas. | — (English madhech vapra) |
| Histogram | Chart of value distribution in bins. | — (English madhech vapra) |
| Hue | Seaborn argument that splits colour by a category. | — (English madhech vapra) |
| iloc | Position-based selection in pandas. | क्रमांकाने निवड |
| IST | India Standard Time (Asia/Kolkata). |
भारतीय प्रमाणवेळ |
| Join / merge | Combining tables on key columns. | की वरून टेबल जोडणे |
| Jupyter | Interactive notebook environment for code + narrative. | — (English madhech vapra) |
| loc | Label-based selection in pandas. | लेबलने निवड |
| Long data | Tidy layout with one row per observation unit (e.g. city-month). | लांब / टायडी स्वरूप |
| matplotlib | Core Python plotting library. | — (English madhech vapra) |
| melt | Transform wide columns into long rows. | रूंद → लांब रूपांतर |
| Missing value (NaN) | Not-a-Number / blank that pandas treats as missing. | गहाळ मूल्य |
| Notebook | .ipynb document of cells (code/markdown). |
— (English madhech vapra) |
| NumPy | Library for numeric arrays; foundation under pandas. | — (English madhech vapra) |
| Outlier | Value far from most others; may be error or real extreme. | टोकाचे / असामान्य मूल्य |
| pandas | Python library for tabular data analysis. | — (English madhech vapra) |
| Pivot table | Cross-tabulated summary with aggregation. | — (English madhech vapra) |
| Reproducible analysis | Same inputs and code regenerate the same results. | पुनरुत्पादनयोग्य विश्लेषण |
| seaborn | Statistical visualisation library built on matplotlib. | — (English madhech vapra) |
| Series | Single column of a DataFrame (or standalone 1-D labelled array). | — (English madhech vapra) |
| Tidy data | Columns = variables, rows = observations — good for analysis. | सुव्यवस्थित / टायडी डेटा |
| Vectorised operation | Computation on whole arrays without Python for-loops. | सदिशित / एकाचवेळी गणना |
| venv | Python virtual environment isolating packages. | — (English madhech vapra) |
| Wide data | Many measure columns (e.g. one column per month). | रूंद स्वरूप |
Ravindra Bagale's Tip
Glossary madhe Marathi arth hard ideas sathi aahe. Interview madhe term English madhech bola — “grain of the table”, “left merge”, “EDA”. Samajla ka?
Thodkyaat sangaycha tar (quick recap)
- Use this page when a chapter term feels fuzzy
- Keep everyday tool words in English
Samajla ka? Abhyas regular theva — Jupyter ughadun roj thoda code. Shubheccha!